A scientific study on the taxation of 'emigrating' capital gains, (Utflyttningsbeskattning av kapitalökningar)

A scientific study on the taxation of 'emigrating' capital gains. (Utflyttningsbeskattning av kapitalökningar.)

by Peter Sundgren
--------------------------------------------------------------------------------

In 1983 Sweden adopted its first emigration tax regime imposing Swedish tax claims on share gains derived by shareholders migrating abroad if the sale of the shares takes place within ten years of the shareholder's departure from Sweden. This rule is therefore called the ten year rule and is embedded in chapter 3 section 19 of the Swedish Income Tax Law.

In an international context legislation of this type is by no means extraordinary. It can be found in many other countries. The rationale for Sweden's tax claim on migrating share gains is manifested (and repeated) e.g. in a government report SOU 2005:99 page 210 when it came up for renewed investigation by the conclusion that the aquisition of the shares may have been financed by loans and that the deductible interest thereon reflects the expectation of taxable annual returns and capital appreciation. Moreover, as far as controlled or closely held share holdings are concerned, dividends and capital gains can be considered to be a deferred taxation of company profits resulting from the labour performed by Swedish resident owners of the company. Originally the ten year rule only applied to the sale of shares issued by Swedish companies (and other specified Swedish securities) but was extended as of 2008 to cover also shares issued by foreign companies (if they had been aquired while the shareholder was resident in Sweden).

It is no exaggeration to say that the ten year rule (together with its interaction with our tax treaties) has become one of the most contentious and controversial rules of the Swedish international tax regime. Considering our relatively high capital (gains) tax of 30 percent and, with regard to closely held company shares, the so called 3:12 rules, which can push the gains into very high marginal income tax brackets, it is no surprise that this taxation has become the target of intense tax planning for shareholders moving abroad. Litigation has been quite intensive, applications for advance rulings very numerous and the number of articles written on this topic has been very large.

In other parts of the world, and one does not need to travel very far, there are a number of tax jurisdictions that levy very low or even non-existent capital gains taxes on share sales. Many of these countries have tax treaties with Sweden conferring the sole right of tax on such gains on the residence country. By moving to such a country and then realising the gain the Swedish tax can thus be avoided. In such an environment where also personal mobility is unrestricted it is tempting for Swedish shareholders to move from Sweden (at least for some time) if they stand to gain a couple of million Swedish crowns. (If also, which is quite common, he is retiring from active occupation, and under a tax treaty can achieve a reduced taxation of his Swedish pension income this of course will also ”sweeten the deal”).

The tax planning techniques that have been adopted in the past to avoid the tax under the ten year rule have been very ingenious and successful mainly, as just mentioned, because our tax treaties confer sole taxing rights on share gains on the residence state i.e. the state to which the shareholder has decided to move. A number of our treaties after lengthy periods of negotiation have, however, secured Swedish source taxation but new holding company strategies and careful planning and timing of transactions have circumvented these obstacles (see below). Since becoming a member of the EU the principles of free movement of individuals and capital have provided new opportunities for further tax avoidance schemes. Also, the abandonment of our foreign exchange rules (in 1989) has played into the hands of migrating shareholders.

The present circumvention technique of the ten year rule is to set up a holding company somewhere in the EU – Cyprus with its low corporate tax rate has become a favourite jurisdiction for this purpose - and then to transfer the Swedish (target) company at aquisition cost to a foreign holding company which can be done tax free under our new participation taxation regime. The shareholder then moves to a (low or no-tax) jurisdiction which as far as capital gains are concerned has a tax treaty with Sweden denying Swedish source taxation of the gains.

The erosion over the years of the Swedish capital gains tax base due to manipulations of the ten year rule has been massive. I repeat, massive. The Tax Administration have reported billions of Swedish crowns escaping the country and repeatedly called for stricter legislation. Consequently the present situation is very lucrative for the international tax planning community1 which together with their clients thus represent a strong force in favour of a status quo.

The Swedish legislator, whatever his political colour, in high-flown rhetoric, has always proclaimed his mission to defend the Swedish tax base. The counteracting measures that have been taken in practice to prevent the abuse of the ten year rule have, however, been lethargic and ineffective. The following example which became a hot issue in 2007 is revealing. It ocurred with respect to Sweden's tax treaty with Austria. In 1992, fifteen years earlier, I repeat fifteen years, Austria had introduced a step-up mechanism regarding capital gains taxation whereby the aqusition price of the shares was tied to their value at the date of immigration of the shareholder to Austria. In this situation, where Austria under the tax treaty with Sweden had sole taxing rights to the gain, an immediate subsequent sale by the shareholder would relieve him from tax not only in Sweden but also in Austria, i.e. a double non-taxation situation. A new treaty had been initialled already in 1995 but still in 2007 – that, ladies and gentlemen, is twelve years later - treaty discussions had not been finalised! According to the Tax Administration the tax loss to the Swedish treasury only in the three years 2004-06 amounted to some three billion Swedish crowns. After quite severe public opinion pressure the Finance Ministry in a whirlwind of activity negotiated a new protocol with Austria closing the tax loop-hole within a record time of about eight months! The legislative measures introduced in 2008 extending the ten year rule to the sale also of foreign shares, the purpose of which was directly aimed at frustrating the tax planning technique decribed above, also represented a total failure as many of our treaties also prohibit source state taxing rights to such capital gains.

The Swedish treasury still generally remains aloof to the tax base erosion caused by the ten year rule blaming staff recruitment problems and calling for further investigation in the matter. That latter proclamation was made in mid 2007 when the ten year rule was extended to foreign shares – a measure which could easily be circumvented. To day, three and a half years later an investigation panel regarding the ten year rule has still not even been appointed.

Recently the Swedish National Audit Office (Riksrevisionen) issued a report criticizing the Treasury for neglecting to take measures against the vast amounts of tax payer's money being lost due to the ten year rule. (RiR 2010:24). Also the Confederation of Swedish Enterprise (Svenskt Näringsliv) has berated the Government for their dilatory treatment of our tax treaty network reducing the competetivness of Swedish companies on their export markets.

In the past I have given a lot of attention to the ten year rule advocating a more energetic approach to counteract the tax avoidance measures that are practised in this field. Considering the very strong forces benefitting from the present situation these efforts have been unsuccessful.

Therefore I found it of great interest in late 2010 to learn that a 500-page doctoral thesis was going to be presented at the Uppsala University regarding emigration taxes. This, I thought hopefully, should invigorate the tax policy debate on the ten year rule in particular. The thesis titled ”Emigration taxation of capital gains – a scientific tax study on international taxation of individuals focusing on problems regarding tax treaties and EU law.” (Utflyttningsbeskattning av kapitalökningar – en skattevetenskaplig studie i internationell personbeskattning med fokus på skatteavtals- och EU-rättsliga problem) is written by Katia Cejie at the Juridical Institution of the Uppsala University and associated also to the private foundation Centrum för Skatter (supported by Deloitte, Ernst & Young, KPMG, Mannheimer Swartling Advokatbyrå, Skeppsbron Skatt och Öhrling Pricewaterhouse Coopers, (all of them of course involved too in international tax planning affairs).

In a press release (www.forskning.se) announcing the upcoming public hearing of her thesis Ms Cejie describes the problems of emigration taxes for one thing as being so considerable that they affect the willingness of individuals to move, in practise restricting the principles of free movement within the EU and, for another, that they under certain circumstances are levied in both the state from which the tax payer is moving and also in his new state of residence i.e. double taxation. ”Ms Cejie's thesis”, says her tutor and Supreme Court judge Kristina Ståhl, ”is important as it provides an instrument to courts and legislators at both the national and EU level for charting the problems in this field of taxation”.

So, how different the description of emigration tax problems can be! On the one hand as a threatening force giving rise to double taxation and limiting the free movement in Europe, and on the other as mentioned above as instruments for wide-spread tax avoidance leading to the payment of quite insignificant taxes or even to complete double non-taxation, serving, if anything, as a driving force for the migration of tax payers!

Ms Cejie's thesis is a very comprehensive, interesting and solid report on various aspects of emigration taxation and it spans a much wider field than just the ten year rule. In particular her research and analysis of EU law in this field and the problems of whether, as aforementioned, emigration taxes constitute restrictions to the freedoms of the Treaty on the Functioning of the European Union (TFEU) and if so, whether or not they can be justified (Part III, Chapters 5-10) have made a big contribution to the understanding of these very complex matters. (She should no doubt be a strong candidate for the post doc award of Centrum för Skatter Foundation.)

Nevertheless, and as shall be gleaned from below, I have a number of serious reservations about Ms Cejie's work mainly on problems arising from the interaction betwen emigration taxes and tax treaties.

The thesis covers four different kinds of emigration taxes (some of which apply also to other types of income than capital gains). This article, however, will focus only on the ten year rule. This is a type of emigration tax categorized in Ms Cejie's thesis as a TL-rule, a trailing tax on latent income which can also be defined as a rule enforcing an exterritorial tax liability (in Sweden, the state from which the shareholder is moving) for a certain period of time. (The term ”trailing tax” is used as a synonym for ”extended income tax liability”, footnote 1716). The definition of the term ”latent income” indicates that taxation is not triggered upon emigration and that the appreciation of the shares that occurs after emigration is also subject to tax (in Sweden) when the shares are sold.

The additional three types of emigration taxes that are dealt with in the thesis are:

a) exit taxes on latent income (EL-rule). A latent income (page 513) is an income which is unrealized, i.e. the ”normal” time of the taxable event has not yet ocurred.The emigration,however, triggers the taxable event which means that that an exit tax is levied upon departure.

b) exit taxes on realised income (ER-rule). One Swedish rule that falls into this category is the rule on exchange of shares. Acording to the Swedish Income Tax Act (ITA), an exchange of shares is a taxable event. The gain is realized but the taxation of the gain is postponed until the new shares are sold or, which is of interest with regard to emigration taxes, until the taxpayer emigrates to a country outside the EEA area. The emigration is, however, regarded as a taxable event according to the rules in this case and an exit tax will be levied.

c) trailing taxes on realized income (TR-rules). One of the Swedish rules that fall into this category is the rule on deferral of the tax levy of the capital gain on the sale of a permanent home. The gain, consequently, is realized but taxation is deferred until the new home is sold. (This rule was introduced in 2007 and applied only to new homes situated in Sweden but the Luxembourg Court of Justice held that this violated the principle of free movement within the Community. This had the effect that also a new home aquired within the Community should qualify for a deferral of the capital gains tax assessed on the sale of one's Swedish home) Ms Cejie concludes that this category of emigration taxes will expand in the future.


Public discussion of Ms Cejie's thesis

Before continuing any further with my analysis of Ms Cejie's thesis I must however report what happened at the public hearing hereof at the University of Uppsala on 10 December 2010.

The chairwoman of the proceedings and Ms Cejie's tutor (and Supreme Administrative Court judge and member of the Board of Centrum för Skatter) Ms Kristina Ståhl set forth the conditions for the hearing. First this would consist of the (cross-)examination by the public opposer Professor Mats Tjerneld, Lund University . Thereafter the certification board (betygsnämnd) consisting of Professor Bertil Wiman, Uppsala University (and Director of the Centrum för Skatter Foundation), Cecile Brokelind, Lund University and Marjaana Helminen, Helsinki University would have the opportunity to raise questions. Finally the floor would be open for comments from the audience. It was explicitly pointed out that there was no specific time frame for the hearing which would thus be concluded only when no more interventions or comments remained.

The two first phases of the proceedings were concluded by approximately 12.20 hours.

I then announced my interest to raise a number of questions. After a short while I was interupted by Ms Ståhl who asked how much more time I required for my interrogation. I responded that this would be difficult to determine because it would depend on the answers and information supplied by Ms Cejie and the discusion that could emerge therefrom. It appeared that Ms Ståhl wanted to know in order to determine if an intermission would be required.

At about 12.40 it was announced that lunch would be served at 13.00 and I sensed that Ms Ståhl was anxious to complete the hearing by then. A break was, however, made at this point during which Ms Ståhl and Professor Mattias Dahlberg (Uppsala University), deputy tutor of the thesis, took me aside for a private talk. They urged me quite abrasively to abstain from delivering any comments to Ms Cejie's thesis and only to ask direct questions to her, everything to speed up the proceedings. I repeated that this disturbed the whole structure of my questioning and that it was necessary for me to develop the various problems, just like the public opposer had done, and also declare my views in order to give Ms Cejie a chance to comment thereon and to explain her opinions in her thesis. I thus repeated that it was impossible for me to determine how much longer time that was necessary for me to conclude my questioning.

At the continuation of the hearing I was again pressured to hurry up by Ms Ståhl. At one point she exclaimed; ”Now that you have put forward your question and received an answer, proceed at once to your next question.” It was thus not possible for me to follow up Ms Cejie's reply which in my view was not quite satisfactory and required further comments.

At this stage, the time was now approximately 12.50, Ms Ståhl declared that that those who so whished could leave for lunch and that she, Ms Cejie, Professor Tjerneld, the certification board and myself would remain to conclude the hearing.

At around 12.55 upon asking another question somewhat ambiguosly answered by Ms Cejie, Ms Ståhl said that I had received a reply and again instructed me to just put forward my next question. At this stage, and with the preparations of my questioning completely disrupted, when I further consulted my notes for a couple of seconds, Ms Ståhl again abrubtly intervened to declare that since I appeared not to have any further questions, the hearing, deespite my protests, was considered to be concluded. And then everybody dropped off for lunch.

Immediately after the hearing I was approached by Professor Wiman who ”thanked me” for having contributed to such a ”pleasant hearing of Ms Cejie's thesis”. Upon my question he confirmed that his statement was ironic. My response hereto was that I considered the pleasantness of a public hearing of a doctoral thesis to be of secondary importance, especially for the members of the certification board. (I have afterwards asked Professor Wiman to confirm his statement but he has not responded hereto.)

After having devoted considerable time to the scrutiny of Ms Cejie's thesis I am frustrated for having been shut up at the hearing and prevented from completing the presentation of my investigation and opinions of her thesis. My complaints hereover have been summarily dismissed by the dean of the juridical faculty at Uppsala University Mr Torbjörn Andersson.

I consider that the basic principle of free and unfettered investigation of all scientific effort, a hallmark of academic tradition of all civilised societies, has been compromised.


Analysis of Ms Cejie's thesis.

The thesis comprises four chapters of which the following discussion will deal only with chapter II sections 2.2 on interpretation of tax treaties and section 2.3. regarding the conflicting norms for determining precedence between domestic tax law and treaties highlighted by the 2008 treaty over-ride ruling of the Supreme Administrative Court RÅ 2008 ref. 24. The following presentation will include not only the issues brought up by me at the hearing but also those that I was prevented to bring up. As already mentioned this report is mainly focused on the ten year rule.

At the hearing I expressed my general disappointment that no attention had been given in the thesis to the historical development of emigration taxation that has taken part in Sweden over the last couple of decades particularly as far as the ten year rule was concerned. As already pointed out, this legislation had constituted one of the most contentious and controversial fields of our international tax regime over a very long time. Nor did the thesis disclose any kind of assessment of the enormous tax base losses suffered by the treasury regarding the ten year rule. Lacking such a historical and empirical background the reader is left unaware of the economic importance of the issues involved in emigration tax matters. The thesis is thus reduced mainly to an academic abstraction isolated from the world outside. Also disappointing is that Ms Cejie had made no attempt to construe a rule that would overcome the shortcomings of the present legislation and of our tax treaties. Her final conclusions and suggestions merely fade away in a deferential and restrained call for further research.


Tax treaty interpretation

Ms Cejie, in chapter II.2.2, has devoted no less than twenty pages to tax treaty interpretation. This, to a certain extent, may be considered unneccessary beacause this, with regard to emigration taxation is of very limited relevance in general and as far as the ten year rule is concerned I cannot recall a single tax case giving rise to any specific treaty interpretation problems. The problem of emigrating share gains is usually addressed in a special section of the capital gains tax article, usually number 13 of our tax treaties and the interpretation of these articles is quite straightforward and uncomplicated.

At the hearing I welcomed, however, the fact that I shared the same basic belief with Ms Cejie that tax treaties should be interpreted according to the principles of international law laid down in the Vienna Convention on the Law of Treaties and that the interpretation of such treaties (only on a very last resort basis) should rely on internal (Swedish) law understanding of terms not defined in the treaty. I noted however that my pioneering work on these matters, a 64 page paper in IUR-INFO no 10-11/1968, had not been noticed by Ms Cejie in her bibliography.

I also brought to attention that an article in IUR-INFO no 2/3 2007 by Stefan Ersson, former judge of the Supreme Administrative Court and a leading authority on international tax law in Sweden, had, in my view, taken a step in the wrong direction when empasizing that tax treaties, being an integral part of Swedish national law, should also be interpreted as such. I had put forward my views in a quite comprehensive response (over 18 pages) rejecting Mr Ersson's ideas suggesting that tax treaties should be interpreted in the same way as domestic tax law. Mr Ersson's article had been quoted by Ms Cejie in her thesis but not my response. The reason herefor, Ms Cejie announced at the hearing, was that my article had not been published. This, in my view, was not a satisfactory answer as it had in fact been printed in ”WebJournal on International Taxation in Sweden” (no 4/2009, Legalitetsprincipen och skatteavtal”) to which Ms Cejie is a subscriber at www.skatter.se Also, the article had been rejected publication in Skattenytt, a decision taken by Professor Mattias Dahlberg who was the editor of that magazine at the time and who, as mentioned above, also has been assistant/deputy tutor to Ms Cejie.

A further issue brought up by Ms Cejie on page 89 of her thesis, discusses the interpretation problems that may occur when treaties have been drawn up in two or more languages both declared equally authoritative. This problem became very real in the Administrative Supreme Court decision RÅ 2004 not. 59 regarding the treaty with Peru which Ms Cejie duly had reported. In 2006 i wrote a 28 page and very critical report, ”Interpretation of tax treaties authenticated in two or more languages – a case study.” on this decision which was published (!) in Svensk Skattetidning 2007. This article too, however, had eluded Ms Cejie's attention.2

A further interpretation issue brought up by Ms Cejie on pages 93-96 is to what extent a change of the Commentaries to the OECD Model Treaty should apply to the interpretation of a bilateral treaty that was made before that change if the text of the pertinent treaty article itself of the Model treaty remains unaltered. This question has been fervently discussed also by Mattias Dahlberg and Jesper Barenfeld in their doctoral thesises and by Anne-Sophie Sallander.3 The main line of their arguments, including also now that of Ms Cejie, is that revised commentaries of the Model Treaty should not be applied or only very carefully to “old” bilateral treaties based on the Model Treaty. Instead one should apply that version of the OECD Commentaries that existed at the time of the conclusion of the bilateral treaty in question, (which of course could be very much earlier).

I have a completely different view on these matters and what I suggest is a useful method for a solution of the problem, which I reported in IUR-INFO 2/2002. and again in a commentary (7 pages) to Jesper Barenfelds thesis, in no 1/2005 of WebJournal on International Taxation in Sweden. This made no impression on Mattias Dahlberg who vigorously defended his position in IUR-INFO 3 under the title “Om rättskällevärdet av OECDs modellavtal jämte kommentar”.

In my view one should remember that the Model Treaty is a document agreed by the the governments of the OECD. These are the same governments that make bilateral treaties with each other. Therefor, if they agree at “the OECD level” that modified Commentaries to the Model should be taken into account also visavi treaties that they have made before the modifications, this reflects an understanding of these governments that they have reached a consensus and given proof of their intention to observe that these treaties shall be interpreted accordingly. A tax treaty is thus a contract between the two contracting states imposing upon one another an obligation to respect and to abide by the rules thereof. In this respect they differ from an ordinary domestic tax law whereby the legislator imposes obligations upon its tax subjects.

In theory, but quite impractical of course, the contracting parties, if they have made a bilateral treaty in the past, the text of which is identical to the Model and if they recognize the validity of the new Commentaries to the Model, could terminate the treaty and then immediately, upon the modifications of the Model Commentaries, let it re-enter into force. In that way there would never be any bilateral model-based treaties that are older than the latest modification of the Model Commentaries!

In my paper regarding Barenfeld's thesis I wrote: Nor is it quite clear why Mr. Barenfeld is so reluctant to the ambulatory approach of using modified Commentaries also to old treaties, but by his reference to the position of the tax payer, it seems that his hesitation stems from a consideration for legal certainty. Or, in other words, the taxpayer must be able to rely on the treaty as it was once concluded. This seemingly suggests that the modifications of the treaty are always to the detriment of the tax payer. But one should of course also remember that the modification just as likely may be to the benefit of the tax payer. In this case everybody would probably agree that it would be a great shame not to apply the modified version of the Model.
At the same time, and this is something upon which everybody agrees, it is very complicated always having to determine at what point a specific modification of the Commentaries took place with regard to the relevant treaty under interpretation. Considering moreover that the Contracting parties as revealed in the recommendations of Model Treaty authors have no reservations about allowing application of modified Commentaries to treaties concluded before these modifications, a useful solution of the problem would be to adopt a method whereby the modified Commentaries shall be accepted for interpretation purposes only with regard to investments made and income derived after the adoption of the modifications. The validity of new Commentaries should thus be referred not to when they were adopted but to the point in time when the income was derived. In this way the problem of certainty for the tax payer is always accommodated. (This is an idea I raised already a couple of years ago, see IUR-INFO 2/02.)

I am pleased to note that this idea has taken root at the Swedish Tax Administration who in a recent opinion (ställningstagande) regarding the Model Commentaries' new thinking on the definition of the term “employer” (article 15) have decided that it should apply as of 2011.

Considering his deep involvment in the issue and in Ms Cejie's thesis it is suprising that Professor Dahlberg has not advised her to include a reference of the above discussion in the thesis. It thus only reflects Dahlberg's opinion on the applicability of new Commentaries to “old” treaties whereas my contribution to this issue has been completely ignored by Ms Cejie.


Purposes and functions of tax treaties (section 2.1.2)

On page 71 of the thesis Ms Cejie brings up one of the purposes of a tax treaty as being to divide taxing rights between the contracting states. This in my view should not be considered as a purpose of a treaty in itself but rather as a means of avoiding double taxation. This is accomplished either by the exemption or the credit method on a reciprocal basis. It is of course a matter of negotiation to determine the level of tax sharing for instance when establishing the level of withholding taxes between the source state and the residence state. Recently Sweden has also concluded a couple of treaties (with Austria and Denmark) where double taxation is avoided not by dividing the tax liability but by splitting the tax base, a rule which, incidentally, applies specifically to capital gains in emigration situations. Sweden as the state of source may thus tax the appreciation of the shares that has taken place up and until the point of emigration whereas the residence state will tax the appreciation that occurs between that point in time and the time of alienation of the shares. The thesis would have benefited from a discussion of this new trend.

A further view regarding the purposes and functions of tax treaties (which there was no time to explore at the public hearing of Ms Cejies thesis) is their role to prevent anti (international) tax avoidance. Ms Cejie's approach and investigation of this matter is very interesting and I agree with her conclusion (pages 121-125) that neither our domestic five year rule imposing (continuing) full tax liability during five years on individuals migrating from Sweden if they have “essential ties” with Sweden, nor our ten year rule imposing capital gains tax liability on company shares on migrating non resident individuals can justify a non-recognition of a treaty just because these Swedish rules are intended to prevent tax avoidance.

A further question regarding tax avoidance and tax treaties with special reference to the ten year rule is to what extent Sweden's general anti tax avoidance regime may be invoked in such situations. This has been highlighted in both the recent Thailand case and the Switzerland and Greece cases (see below) focussing basically on whether the ten year (emigration tax) rule should override Sweden's tax treaties with these countries. The bottom line decision by the Advance Rulings Board in the first mentioned case was that the treaty prevailed but in the two latter cases the Board in accordance with the Supreme Administrative Court's ruling in RÅ 2008 ref.24 gave the opposite answer i.e. that the ten year rule had precedence over the treaties, i.e treaty override. In these two latter cases three members of the minority addressed the question of the relevance of the general anti-avoidance rule visavi the treaties (all of them coming to the conclusion that it did not apply as such.)4

I have written a comprehensive report on the Thailand case (see below) which, after it too was denied publication in Skattenytt by its editor Professor Dahlberg, was printed in no 4/2009 of WebJournal on International Taxation in Sweden. It has, however, not been considered by Ms Cejie.


The “golden rule” and conflicting tax liabilities under domestic and treaty law.

This is a very big question in general and because of Ms Cejie's opinion one of some controversy visavi my standpoint. Due to the limited time allowed for my intervention at the hearing this was discussed only briefly.

The golden rule is laid down, (usually in paragraph 2), in each and every law incorporating our tax treaties. It reads as follows: “The rules of the treaty shall be applied only to the extent that they give rise to a limitation of the tax liability that would otherwise occur.” The term “that would otherwise occur” naturally refers to the tax liability that would follow if no treaty existed i.e. the tax liability resulting from application of pure domestic law. The application of the golden rule is actually quite straightforward: First one determines how much tax Sweden can levy under domestic tax law. If no taxation occurs the matter is closed. There is no need to consult the treaty. If, on the other hand a tax liability occurs under Swedish tax law, then one must consult the appropriate treaty and determine how much tax should be paid in Sweden according to the treaty. If this amount is lower than that which follows from the application of Swedish law, Sweden may not levy a higher tax than what follows from the application of the treaty. Important to note is that the golden rule is nothing that follows from some kind of principle or general (international) maxim. It is a Swedish (unilateral) decree/law imposed by our legislator (parliament). It is repeated in each and every law incorporating our treaties. These laws are hierarchically equal to any other law enacted by parliament including our domestic tax laws and our tax treaties. None of these laws are superiour to any of the others.

On page 75 of her thesis Ms Cejie, referring inter alia to an article I wrote in 1995 (page 463), points out: “A common opinion in doctrine is that treaties are superior (emphasis added) to domestic law and that this is based on the so called golden rule. I don't share this opinion about the golden rule but I raise the question in this context based on the said opinion”. After having quoted the golden rule and given an explanation of its limiting effect as reported above Ms Cejie concludes that “if on the other hand the treaty will allow Sweden to tax so be it.” This must be understood that there is no limiting effect by the treaty and Sweden can go ahead and levy the tax that follows from an application of its domestic law. On pages 76-77 Ms Cejie then continues: “The golden rule has further been been interpreted as conferring upon treaties a superior role visavi domestic law. Also in doctrine abroad the golden rule has been interpreted as giving preference to treaties where a conflict of laws appears. This approach, that tax treaties have preference to domestic law,has been held for a long time. I am , however, not convinced that this viewpoint can be derived from the golden rule. According to its wording the golden rule says nothing about the sovereignty (dignitet) of the various legal sources (i.e. that tax treaties are superior to domestic law, a form of vertical effect). Thus I do not share the customary view in doctrine about the understanding of the golden rule. In my opinion the golden rule should be understood as nothing more than a rule under which the tax liability under domestic law cannot be extended by a treaty (a form of horisontal effect). This means that if Sweden is afforded taxing rights under the treaty this taxing right must be confirmed also under national law”.

Sure enough, Sweden can of course never impose taxation, treaty or no treaty, unless confirmed in national Swedish law. Or in other words, we don't need a treaty to confirm our right to tax under national law! It is also true that neither the golden rule nor a treaty rule says anything about the superiority (or inferiority) of either rule or domestic law. Under our constitution on the other hand domestic law, treaty law and the golden rule itself, which is indeed also a domestic law rule adopted by parliament, all have the same constitutional “ranking”. None of these rules therefore have preference as such before any ot the others. As already mentioned none of them is superiour to the others from a constitutional viewpoint. And nothing else, in Swedish doctrine, has ever been suggested! Ms Cejie's allegation (above) that traditional doctrine views, quoting my 1995 (page 463) article in Skattenytt and Bokström,K&Tyllström's national report in Cahier de Droit Fiscal 1993, that treaties are superior to domestic laws represents a serious misunderstanding. As I pointed out in my 1995 article, the role of the golden rule is to be nothing but a tie-breaker (lagvalsregel) when tax liability under Swedish tax law differs from that of the treaty. It is true, as Ms Cejie suggests, that tax treaties cannot extend Swedish tax claims under Swedish law. But that is only bacause the golden rule explicitly says so. If our parliament so chooses it could alter the golden rule so that treaties always have precedence. Nor would such a change of the golden rule change the equilibrium of the ranking between domestic and treaty law. The limiting effect of treaties i.e. that in such situations they trump domestic law, and I repeat it once again, is nothing that follows from nature. It is, as just explained, due to an act of parliament.

Professor Mats Tjerneld, the public opposer of the thesis, brought Ms Cejie's opinions on the golden rule to attention contenting himself, however, and to my disappointment, by merely establishing that they were bold and interesting.

I have further expanded my views regarding the relationship between domestic law and tax treaties in an article under the title “Montesquieu and the Swedish income tax regime (and some thoughts on the relationship between domestic tax law and tax treaties.)”. It was printed in “WebJournal on International Taxation no 6/2007 after being denied publication in Svensk Skattetidning for being “too critical”. Nor has this article been cited in Ms Cejie's thesis.


The OMX-case RÅ 2008 ref. 24

This case presented by the Supreme Administrative Court on 3 April 2008 deals specifically with the problems discussed in the preceding section regarding the relationship between domestic tax law and tax treaties and represents probably the most criticized decision ever by that Court, certainly with regard to tax treaty matters. Ms Cejie discusses the case over no less than fifteen pages under sections 2.3. and 2.4. of her thesis.
In footnote 327, under what appears to be an exhaustive list, Ms Cejie has quoted eleven authors who have discussed the OMX case. In addition to her list I have written another three articles in various contexts (in Swedish) on the subject which, however, have either not been observed by Ms Cejie or have been ignored by her. They are published in:
1.No 4/2008 of WebJournal on International Taxation in Sweden, “Treaty Override”, http://www.skatter.se/index.php?q=node/2533.
2.No 5/2009 of WebJournal on International Taxation in Sweden, “Försäljning av aktier efter utflyttning, del 1 (Thailandsmålet) more hereinafter, http://www.skatter.se/index.php?q=node/2625.
3 No 3/2010 of WebJournal on International Taxation in Sweden, “Normhierarki och regelkonkurrens mellan intern skatterätt och skatteavtal”, http://www.skatter.se/index.php?q=node/3318.

The OMX case dealt with the question whether our Controlled Foreign Corporation (CFC) legislation should prevail over our tax treaties, (a question which has arisen in several other countries with CFC-legislations). The Advance rulings Board, after a very interesting and meticulous analysis and interpretation of the pertinent tax treaty (with Switzerland), found that this was the case i.e. that Sweden could go ahead and tax and that the treaty did not prevent Sweden from executing its domestic CFC-rules. On appeal the Supreme Administrative Court came to the same bottom line conclusion but they did so without consulting the treaty at all! They explicitly declared that there was no need to analyze the treaty.

Or as professor Mutén put it, the Court resolved the matter “by a Gordian stroke of the sword” allowing the domestic Swedish CFC rules to override the treaty. This is what – quite majestically - the Court said: “Upon making a tax treaty Sweden (may) surrender tax claims following from Swedish law. From an international law point of view Sweden is bound by the treaty. The obligations under the treaty become effective by adoption in Swedish law. As far as Switzerland is concerned this took place according to Law (1987:1182) on (incorporation of) the tax treaty between Sweden and Switzerland. Such a law has no exeptional ranking in relation to other laws. When thus, as outlined under paragraph 2 of of the said law in its version under Law 1992:856, it is declared that “the rules of the treaty shall apply only to the extent that they incur a limitation of the tax liability in Sweden that would otherwise occur” this only suggests that the treaty cannot expand the tax claims determined from Swedish law. Therefore, the stipulation does not prevent Sweden in a subsequent law to expand its tax claims regardless of the repercussions that would entail under international law. If two laws prove to be irreconcilable the dispute over which law should prevail shall be resolved according to the principles of conflicts of laws. In doing so it is clear that not only have the CFC rules been adopted after the incorporation of the treaty but also that they take aim specifically at such activities as performed by the Swiss company. Accordingly it is clear that the CFC rules shall prevail and have effect regardless of the result following from the treaty rules. An analysis of of the regulations of the treaty is thus not required.”

The ruling is very cryptic and the number of arguments that have been suggested analyzing the case are innumerable. The first article that was put forward was by professor Mattias Dahlberg of Uppsala University, and one of Sweden's leading authorities on international taxation in Sweden. The article was published in the summer of 2008 in Skattenytt (of which Mr Dahlberg was then chief editor) 2008 pages 482-489. His main line of thought is that the Court had abandoned its traditional principles under international law for the interpretation of tax treaties. In no less than sixteen different contexts of his article Mr Dahlberg refers to the erroneous interpretation of the treaty committed by the Court suggesting further that it had adopted a domestic principle of tax interpretation in conflict with the international law method (folkrättslig metod) laid down by the Court in its (famous) “Luxembourg ruling” RÅ 1996 ref. 84 under which the interpretation of tax treaties should be governed by the international law standards of the Vienna Convention on the Law of Treaties as they appear in articles 31-33.

In response to Professor Dahlberg's article I also slated the treaty override perpetrated by the Court in RÅ 2008 ref.24 but strongly disagreed with Dahlberg's suggestion that this had its root in an interpretation of the treaty. As already mentioned the Court had completely ignored the treaty, very clearly determining that it did not require any analysis. And an interpretation of a treaty can only follow from an examination or an analysis of the treaty text. Or in other words, how can one interpret a treaty text if it is not analyzed? (And how could the Court have reached the conclusion that the CFC rules constituted “lex specialis” visavi the treaty if the treaty had not been considered?)

My paper that was sixteen pages long was submitted for publication in Skattenytt in October 2008. But it took Professor Dahlberg only eighteen minutes (according to the time indicated on his email) to return it, rejecting publication on the grounds that it added nothing of substance to the matter.

My analysis of the case which I further explored in my article (3 above) titled “Hierarchy of norms and conflicts of law between domestic tax laws and treaty rules” and which contrasted (I believe) to all of those of other authors was that the conflict of law suggested by the Court was not between the domestic CFC rule and the treaty but between the CFC rule and paragraph 2 of the law of incorporation of the treaty. This article was also denied publication, not by Professor Dahlberg but by his colleague at the Uppsala University Assistant Professor Jan Bjuvberg who is editor of Svensk Skattetidning but is, as mentioned above, printed in the WebJournal on International Taxation in Sweden.

Ms Cejie's opinion on the OMX-case, apart of course for concluding too that it constitutes a treaty override, is not alltogether clear. She has not dismissed the opinion of Professor Dahlberg, her assistant tutor. On the other hand, (page 105) she concludes that it appears from the wording of the judgement that no interpretation process has taken place. In footnote 338 she also concurs with the opinion of Maria Hilling (Intertax, Vol.36, Issue 10, 2008) who (page 458) comes to the conclusion too that the Court did not find it necessary to interpret the treaty in order to deal with the question of whether or not it hinders CFC taxation. I found it quite frustrating not to have been given the opportunity to discuss this matter at the public hearing of Ms Cejie's thesis.

It is of course also disappointing to have one's papers denied publication by the most prominent tax periodicals but as such decisions also may be based on commercial grounds this is something that every author must accept. In an academic context, however, ignoring, as Ms Cejie has done, what in this case is relevant material is a different matter. It affects the scientific quality of her thesis.


The Thailand case. (Decision 73-08/D, April 2 2009 by the Advance Rulings Board)

This case also deals directly with the aforementioned problem regarding the relationship between domestic and treaty law and the disastrous treaty override result in the Supreme Administrative Court's decision iRÅ 2008 ref.24. Also of interest is that the Thailand case is all about emigration taxation dealing specifically with the priority between the domestic ten year rule and the capital gains article of, in this case, the Thailand) treaty.

This was the situation that arose in this case. A Swedish resident shareholder of a Swedish company formed a Netherlands company in1998. In 2008 he transferred his Swedish company to the Netherlands company (taxfree under Swedish law). He then moved to Thailand and sold the Netherlands company. He applied to the Swedish Advance Rulings Board wondering where the gain should be taxed? Under the Swedish ten year rule which had been changed in 2008 and extended to cover also capital gains on all kinds of foreign shares sold within ten years of emigration from Sweden, it was clear that Sweden could tax the gain on the sale of the Netherlands company. Under the treaty with Thailand made in 1998 it was, however, equally clear that Thailand had sole taxing rights. The Advance Rulings Board exempted the shareholder from Swedish tax according to the treaty. The great surprise in this case was that the Board ignored the Supreme Administrative Court's decision RÅ 2008. ref 24. under which à "new" domestic rule should trump an old treaty.

There has been some controversy over the interpretation of this case which focuses on the basic principles of the relationship between domestic law and treaty law.

Professor Dahlberg who is (also) a judge of the Advance Rulings Board and participated in the Thailand ruling declared that there was no conflict between national law and the treaty, thereby indicating that under the ten year rule too the gain was exempted from Swedish tax. Or in other words, both domestic law and the treaty exempted the gain from Swedish tax. But if so, as I suggested in my article, the Board would not have exempted the gain according to the treaty as mentioned above but according to domestic law. Or, in other words, if domestic law exempts an income from taxation there is no reason to consult a treaty to establish the same exemption. In my article I therefore pointed out professor Dahlberg's misunderstanding of the case. As already mentioned, this article was denied publication in Skattenytt, of which magazine professor Dahlberg at the time was editor, but subsequently printed in ”WebJournal on International Taxation in Sweden”, issue no 5/2009 ,”Försälning av aktier efter utflyttning, Del 1, (Thailandsmålet). See http://www.skatter.se/index.php?q=node/2625. (The article was also denied publication by Svensk Skattetidning, a tax periodical of which' editorial board Ms Kristina Ståhl is one of the members).

Ms Ann-Sofi Sallander of The Jönköping International Business School, who's article ”I kölvattnet av RÅ 2008 ref. 24” was indeed published in Svensk Skattetidning also addresses the Thailand case repeats Professor Dahlbergs mistaken opinion. Or in other words, she also considers that the capital gain was exempted from tax also under domestic Swedish tax law.

This controversy, of course, made me very curious to hear Ms Cejie's view on the matter because there was no mention thereof in her thesis. At the public hearing, on my direct question, she agreed with my interpretation of the case i.e. that there was indeed a conflict between the ten year rule and the Thailand treaty and that the treaty had prevailed. Considering that Professor Dahlberg was also her assistant tutor one would have expected that she would have mentioned that she disagreed with his ruling. Nor has Ms Cejie in her thesis mentioned or analyzed the important fact that the Advance Rulings Board in their decision of the Thailand case, ignored the Supreme Administrative Court's treaty override in RA 2008 ref 28. Also she has ignored or overlooked my aforementioned article. This, in my view, and as already suggested does not live up to acceptable standards of academic research.


Emigration taxation and double taxation

In her thesis Ms Cejie has expressed repeated concern over the fact that emigration taxes so often may give rise to double taxation. Also in tax treaty situations. The reason herefor is simply that tax treaties, as mentioned by Ms Cejie on page 218, are not written and negotiated with emigration taxes in mind. The only remedy in most of these cases is to appeal to the competent authorities of the tax treaty for a mutual agreement settlement by the contracting states. This, as also pointed out by Ms Cejie, involves, however, a beaureaucratic procedure and an unpredictable outcome.

The ten year rule will, however, never give rise to double taxation if a tax treaty exists. The treaty will either give sole taxing rights to the residence state excluding completely the source state (Sweden) from taxation or it will give Sweden primary right of tax with an obligation for the residence state, i.e. the state to which the shareholder has moved, to allow a credit for the Swedish tax or, more seldom, to exempt the gain. The problem with this latter type of treaty is for Sweden to negotiate it because it means that the other state must pick up the tab to avoid double taxation. This may then oblige Sweden to make other sacrifices in the treaty either by reducing the scope of the ten year rule normally by reducing the ten year ”quarentine” period or give up other taxing rights.

At the public hearing , and seemingly with Ms Cejie's approval, I broached the idea that Sweden in these situations should offer a reverse credit mechanism i.e. allowing a credit for the capital gains tax of the residence state tax. Considering that the tax avoidance strategies used regarding the ten year rule will always involve low or no tax jurisdictions the tax credit will normally be quite negligible or even non-existent and thus very inexpensive for the Swedish treasury and probably frustrate the tax planning alltogether.

(Sweden has adopted reverse credits in some of its tax treaties regarding Swedish source pension income. As such tax is levied in Sweden as a preliminary (monthly) source tax upon payment and can be reclaimed only after a final tax assessment in the residence state has taken place quite some time thereafter, the reverse credit mechanism gives rise to a temporary double taxation and a complicated administrative procedure and has thus been justifiably crticized. These problems are, however, not relevant with regard to capital gains taxes.)

A further and more sweeping suggestion for reducing double tax problems involving the ten year rule would simply be to limit the scope thereof, at least as far as treaty states are concerned. One could thus limit its range only to closely held or controlled companies and exempt gains below certain thresholds. And in our treaties one could suggest to exempt citizens of the other contracting state from the ten year rule tax.

The double taxation problem as far as emigration taxes are concerned are, as just pointed out, very thoroughly discussed by Ms Cejie. This topic is thus repeatedly dealt with separately for each and every category of emigration tax both EL-, ER-, TL- and TR-rules under the subtitle ”Methods for elimination of double taxation” and then again under a concluding section. The other ”side of the coin”, i.e. the situation where emigration taxes, particularly the ten year rule tax, give rise to unwarranted double non-taxation effects is, however, completely ignored. Considering the severe effects of this aspect of emigration taxation on capital gains a discussion thereof in the thesis would have been justified.


Tax collection problems regarding trailing taxes om realised income (TR).

The statement in one of my opening remarks of the lack in Ms Cejie's thesis of any references to empirical data on emigration taxation is not alltogether correct. On page 507, in one of the closing sections of her work (11.5.4. Conclusions and viewpoints for the future) she gives some figures on the amounts of tax deferred with regard to trailing taxes on realised income (TR-rules). A big problem with such rules, she says, "is the difficulty, despite the (EU) directive on tax collection, to determine and supervise when to exercise the tax levy. The deferral of the tax can easily become a permanent exemption unless the collection of tax is effectivly monitored”. Also, she observes: ”The accumulated total amount of deferred capital gains taxes regarding private home dwellings in 2008 amounted to 216 billion SEK corresponding to a tax claim of 50 billion. The number of deferrals regarding exchanges of dwellings in the 2000-2008 period rose by 600 percent. Of the 216 billion only 1.3 billion apply to deferrals of home owners who have moved abroad. With regard to share exchanges that qualify for tax deferral the amount of capital gains taxes deferred amounted to about 15 billion SEK out of which 257 million SEK applied to shareholders that had moved abroad. The (aggregate) amount of deferrals concerning capital gains taxes of tax payers that have moved abroad amounted to about 1.5 billion SEK corresponding to a potential tax shortfall (emphasis added) of 354 million SEK." In a footnote Ms Cejie then also observes that "despite this high amount it only represents 0.03 percent of the total amount of tax collected by the treasury". Ms Cejie thus starts out by pointing out that the potential shortfall of tax attributable to foreign capital gains represents "a big problem" but finally concludes that it amounts only a fraction of a percent of total taxes due!

A more unsettling aspect is, however, that Ms Cejie in this context has completly failed to mention the shortfall of tax attributable to share gains avoided by manipulations of the ten year rule. As already noted these amounts are quite enormous, and has prompted the Tax Administration to call for emergency measures. In their official report no 1 of 2008 of the combined tax shortfall in Sweden (Skattefelskarta för Sverige) the amount of tax avoided under the ten year rule rises to about 1.2 billion SEK or about four times the figure attributable to deferrals of gains relating to foreign home owners and share exchanges quoted by Ms Cejie. Moreover, the 1.2 billion is an annual figure whereas the other taxes represent an accumulated amount.

Ms Cejie's opinion regarding the aforementioned potential tax shortfalls due to the problems of tax collection under TR-rules echoes the same arguments raised in an official government investigation (SOU 2009:33) regarding the ten year rule conducted by County Administrative Court judge Leif Gäwerth. As I pointed out in a quite disparaging analysis of his report, I suggested that an argument advocating non-interference with the present and aforementioned factual tax shortfall resulting from the manipulation of the ten year rule compared to a potential risk of non-collection under prospective new measures against such abuse was, with all due respect, quite silly. Or in other words, the risk of a potential tax shortfall under a TR regime compared to the present TR situation where taxation can easily be avoided is not a very strong argument for rejecting a change of the rules. Moreover, and more importantly, Mr Gäwerth, and consequently also Ms Cejie, have completely ignored the preventive or obstructive effects of a TL-rule on emigrating share gains. The manipulations of the ten year rule, setting up holding companies in various countries and intra-group company transfers only for tax avoidance purposes, has given rise to a virtual ”industry” in international tax planning circles. By introducing a TL-rule on share gains – something that has successfully been done in other countries – where such gains eventually will trigger a taxation in Sweden would thus frustrate such tax planning alltogether. Or in other words, a market where a tax planning/avoidance scheme only defers the tax temporarily will quickly dry up. Finally, which I also pointed out in my article regarding Mr Gäwerth's report, one should consider the growing effectiveness of exchange of information and tax collection agreements and procedures between treaty partners, including the growing numbers of such treaties also with notorious tax haven jurisdictions, which no doubt should discourage tax payers from not reporting their taxes.

My article regarding Mr Gäwerth's report was first approved for publication in Svensk Skattetidning by its editor, the aforementioned Jan Bjuvberg, but this decision was then abruptly revoked by himself (!) and the editorial board without explanation.

The article was subsequently printed in no 1/2010 of WebJournal on International Taxation in Sweden but, seemingly, has been overlooked by Ms Cejie.


Conclusion

The above observations regarding Ms Cejie's thesis should not overshadow the valuable and interesting work that she has done regarding emigration taxation. I repeat, however, my regret of the fact that her thesis has been isolated from its historical context in Sweden and the socioeconomic realities of the world outside. Scientific research must never be confined to a closed academic 'ivory tower'.

Emigration taxes, and particularly the ten year rule, give rise to considerable problems in international taxation in our country. Not only those of double taxation and freedom of movement within the EU which have been the main topics of Ms Cejie's thesis. But also, as especially share gains are concerned, they lead to unbridled tax base erosion and therefore also to an erosion of the integrity of the tax system as such. Other countries have suffered from similar problems but have taken (as it seems) effective counteracting measures. This, also, would have merited a discussion in Ms Cejie's thesis. At the public hearing Ms Cejie responded affirmitively to my question whether she thought she could construe effective rules that would both prevent double taxation and promote free movement within the EU but also restrain the widespread tax avoidance resulting from the manipulation of emigration taxes on capital gains. Also, the Commission has – quite some time ago - provided guidelines for member States how to deal with these problems, Kom (2006) 825 final. These aspects inter alia have also been dealt with in an excellent study over 53 pages by Anna Buhre and Elisabeth Mörck at the Jönköping International Business School under the title ”Is the ten-year- rule in chapter 3 section 19 of the Income Tax Law in compliance with EC-law?, a study which also has been overlooked by Ms Cejie.

What a pity therefore that Ms Cejie didn't take the opportunity to ”rock the boat” in this field of taxation! With her background, knowledge and experience regarding emigration taxes her voice would no doubt have been heard in shaping public opinion in this regard, prompting the Swedish legislator to get moving. Instead he can now go back to sleep and migrating Swedish shareholders and their tax advisors (including the benefactors of Centrum för Skatter) can relax and keep on laughing all the way to the bank.....

Stockholm May 2011

peter@sundgren.net