Experimental feature warning
The hosting of judicial decisions on Tax Interpretations is in Beta. We do not guarantee the accuracy of this text.
The text-formatting is also in development. The final version will be more polished than what you see here.
Federal Court Reports
Silicon Graphics Ltd. v. Canada (C.A.)  1 F.C. 447
Neutral citation: 2002 FCA 260
CORAM: STONE J.A.
SILICON GRAPHICS LIMITED
- and -
HER MAJESTY THE QUEEN
Heard at Toronto, Ontario, on Monday, May 13, 2002.
Judgment delivered at Ottawa, Ontario, on June 17, 2002.
REASONS FOR JUDGMENT BY: SEXTON J.A.
CONCURRED IN BY: STONE J.A.
Neutral citation: 2002 FCA 260
CORAM: STONE J.A.
SILICON GRAPHICS LIMITED
- and -
HER MAJESTY THE QUEEN
REASONS FOR JUDGMENT
 The issue in this case is whether a company is controlled by non-residents for the purposes of paragraph 125(7)(a) of the Income Tax Act, R.S.C. 1985, c. 1 (5th Supp.), by reason solely of the fact that more than 50% of the shares are held by non-resident persons where there is no evidence of any common connection among them.
 The appellant is the successor corporation to Alias Research Inc. ("Alias"). Alias was incorporated in 1985 under the Ontario Business Corporations Act, R.S.O. 1990, c. B-16, and was at all material times in the business of creating and marketing advanced computer graphics software products. From 1986 to 1993 Alias employed a significant and increasing number of people doing scientific research and experimental development ("SR & ED") in Canada. Alias's principal place of business was in Toronto, Ontario.
 From February 13, 1985 until July 17, 1990, Alias was not a publicly traded corporation. The majority of outstanding shares of Alias during this time were held by Canadian residents.
In those years Alias was recognized by Revenue Canada as a "Canadian-controlled private corporation" ("CCPC") as defined in the Income Tax Act. It was therefore eligible for investment tax credits on its qualified SR & ED expenditures at the rate of 35% on the first $2,000,000 of such expenditures pursuant to subsection 127(10.1) of the Income Tax Act and for refundable income tax credits under subsection 127.1(1) of the Income Tax Act.
 On July 17, 1990, during Alias's 1991 taxation year, Alias made an initial public offering ("IPO") of common shares through the NASDAQ stock exchange in the United States. A total of 5, 049,836 common shares were issued following the public offering. No shares other than common shares were outstanding after the IPO.
 Alias subsequently issued common shares as a result of private placements, exercise of employee share options, consideration paid for corporate acquisitions and fees for services rendered by third parties, from November of 1990 through the end of Alias's 1993 taxation year. By January 31, 1993, there were approximately 8,187,241 issued and outstanding common shares of Alias.
 Following the IPO and thereafter, including to the end of Alias's 1993 taxation year, more than half of the common shares were held by non-residents of Canada. At the end of the 1992 and 1993 taxation years, non-residents held 89% and 74% of the common shares respectively.
 It should be noted, as well, that Alias was not only publicly traded but relatively widely held. During the relevant times no shareholder held more than 13% of the shares. On September 10, 1991, there were 136 shareholders, 78 of whom were non-residents. On May 6, 1992, there were 305 shareholders, 233 of whom were non-resident.
 There was no evidence of any agreement or common connection among the shareholders that would influence the manner in which they would vote their shares and indeed it appears there was no easily available mechanism for shareholders to learn the identity of the other shareholders.
 During this time, the majority of the board of directors and the entire management team were residents of Canada and Alias's principal place of business was in Toronto, Ontario.
 The Toronto management team annually prepared a slate of people to be elected to the board of directors, which slate was always accepted by the shareholders.
 In assessing Alias for its 1992 and 1993 taxation years, the Minister concluded that, because more than 50% of the shareholders of Alias were non-resident, Alias no longer met the statutory definition of a CCPC. As a result, the claims by Alias in respect of SR & ED were disallowed.
Decision of the Tax Court
 The issue before the Tax Court was whether Alias was a CCPC. A CCPC was defined in subsection 125(7) of the Income Tax Act which at the relevant times read as follows:
"Canadian-controlled private corporation" means a private corporation that is a Canadian corporation other than a corporation controlled, directly or indirectly in any manner whatever, by one or more non-resident persons, by one or more public corporations (other than a prescribed venture capital corporation) or by any combination thereof.
Insofar as it applies to the present case, this definition breaks down into three elements: to be a CCPC, a corporation (1) must be a Canadian corporation, (2) must be a private corporation, and (3) must not be controlled by one or more non-resident persons.
 A "Canadian corporation" was defined in subsection 89(1) of the Income Tax Act as follows:
"Canadian corporation" at any time means a corporation that was resident in Canada at that time and was ... incorporated in Canada, ...
Alias was incorporated in Canada and was resident in Canada and therefore was a "Canadian corporation".
 The definition of a "private corporation" in the Income Tax Act as contained in sub-paragraph 89(1)(f)(i) provided:
"private corporation" at any particular time means a corporation that, at the particular time, was resident in Canada, was not a public corporation and was not controlled by one or more public corporations ... or prescribed federal Crown corporations or by any combination thereof ...
 In the relevant taxation years, "public corporation" was restrictively defined in paragraph 89(1)(g) of the Income Tax Act:
"public corporation" at any particular time means a corporation that was resident in Canada at the particular time, if ... at the particular time, a class or classes of shares of the capital stock of the corporation were listed on a prescribed stock exchange in Canada ...
The relevant portion of the definition provided that a "public corporation" means a corporation which has a class of shares "listed on a prescribed stock exchange in Canada". Section 3200 of the Income Tax Regulations, C.R.C., c. 945, lists which exchanges are "prescribed" for the purposes of section 89. The NASDAQ was not a stock exchange in Canada and, during the relevant taxation period, was not a prescribed stock exchange under Section 3200. Therefore, subsequent to the IPO, Alias was not a "public corporation", given the manner in which "private corporation" was defined and, as it was resident in Canada, Alias was, by default, a "private corporation" and a "Canadian corporation". Therefore, the sole issue before the Tax Court Judge was whether Alias was "controlled directly or indirectly in any manner whatever by one or more non-resident persons".
 The Tax Court Judge dismissed the appeal and held that de jure control existed by reason of the simple fact that a majority of the outstanding shares of Alias were held by non-residents. As a result, the Tax Court Judge found it unnecessary to consider the issue of de facto control raised by the respondent in its pleading.
 The Tax Court Judge concluded:
Once the number of non-resident shareholders reaches 50 percent plus one, the control and right to elect the Board of Directors has passed to those non-resident shareholders and a common connection between those non-resident shareholders is not a requirement.
Issues on Appeal
 The main issue is whether Alias was a CCPC in its 1992 and 1993 taxation years. Specifically, the issue is whether Alias was controlled by non-residents during its 1992 and 1993 taxation years.
 This main issue breaks down into two sub-issues:
1) Was Alias subject to de jure control by non-residents?
2) Was Alias subject to de facto control by non-residents?
De Jure Control
 Where the issue of the definition of "control" has arisen in the past, a distinction has been made between de jure control and de facto control. These will each be considered in turn.
 The general test for de jure control has been described in a number of cases as the majority voting control of the corporation, as manifested by the ability to elect the directors of the corporation. The classic description of de jure control was articulated by Jackett P. in Buckerfield's Ltd. v. Minister of National Revenue (1964), 64 DTC 5301 at 5303, where he defined de jure control as,
the right of control that rests in ownership of such a number of shares as carries with it the right to a majority of the votes in the election of the board of directors.
 This statement has been quoted with approval by the Supreme Court of Canada in a succession of cases: M.N.R. v. Dworkin Furs (Pembroke) Ltd. (1967), 67 DTC 5035 at 5036;  S.C.R. 193">Vina-Rug, (Canada) Limited v. M.N.R.,  S.C.R. 193 at 197; International Iron & Metal Company Limited v.  S.C.R. 193">M.N.R. (1972), 72 DTC 6205 at 6207; The Queen v. Imperial General Properties Limited (1985), 85 DTC 5500 at 5502; and Duha Printers (Western) Ltd. v. Canada,  1 S.C.R. 795 at 815. It should be noted, however, that in none of these cases did the Supreme Court of Canada determine that de jure control comprised a simple majority of shares in a widely-held company. In every case, the controlling interest was owned by one shareholder or a handful of shareholders that were connected in some way.
 The most recent pronouncement by the Supreme Court of Canada on the concept of de jure control is that contained in Duha Printers, supra, relied upon heavily by the respondent. In that case, Duha Printers (Western) Ltd. ("Duha") decided to acquire the shares of an inactive company, Outdoor Leisureland of Manitoba, from Marr's Leisure Holdings in order to take advantage of Outdoor's non-capital losses.
 Subsection 111(5) of the Income Tax Act restricted the claiming of losses by a corporation in circumstances where "control of the corporation has been acquired by a person or group of persons" that did not control the corporation when the losses were incurred. Therefore, in order for Duha to deduct Outdoor's non-capital losses, Duha had to purchase Outdoor and yet Marr's had to retain control of Outdoor. To accomplish this, it was arranged that Marr's would buy a controlling interest in Duha and then Duha would buy the shares of Outdoor.
 The issue in Duha Printers was whether Marr's acquisition of 56% of the shares of Duha, which Marr's held for a period of one day, amounted to acquisition of control of Duha. The Minister argued that Marr's brief ownership of Duha meant that Marr's did not control Duha. Nevertheless, the Supreme Court found that Marr's had acquired control of Duha (and therefore Outdoor) for the purposes of the Income Tax Act.
 Dealing with the issue of control of a corporation, Iacobucci J. in Duha Printers said at 815:
Thus, de jure control has emerged as the Canadian standard, with the test for such control generally accepted to be whether the controlling party enjoys, by virtue of its shareholdings, the ability to elect the majority of the board of directors. However, it must be recognized at the outset that this test is really an attempt to ascertain who is in effective control of the affairs and fortunes of the corporation. That is, although the directors generally have, by operation of the corporate law statute governing the corporation, the formal right to direct the management of the corporation, the majority shareholder enjoys the indirect exercise of this control through his or her ability to elect the board of directors. Thus, it is in reality the majority shareholder, not the directors per se, who is in effective control of the corporation. This was expressly recognized by Jackett P. when setting out the test in Buckerfield's.
And he further said at page 817:
The general approach to the determination of control, as I have already noted, has been to examine the share register of the corporation to ascertain which shareholder, if any, possesses the ability to elect a majority of the board of directors and, therefore, has the type of power contemplated by the Buckerfield's test, supra.
 The respondent argued that Duha Printers supported its position that one should only look at the share register in order to learn whether non-residents had control of Alias and that, if more than half of the shares were held by non-residents, then the non-residents had control. In particular, the respondent relied on the following statement by Iacobucci J. at 827:
[A]greements among shareholders, voting agreements, and the like are, as a general matter, arrangements that are not examined by courts to ascertain control. In my view, this is because they give rise to obligations that are contractual and not legal or constitutional in nature.
 The respondent submitted that this statement suggests that any other factual nexus or connection between individual shareholders is not to be examined for the purpose of ascertaining whether or not de jure control is held by those shareholders.
 It must be remembered that in Duha Printers, the Court had before it a case in which a single shareholder had a majority of shares in the order of 56%, which normally is sufficient to demonstrate de jure control. The main issue in Duha Printers was whether or not a unanimous shareholder agreement (USA) could be considered in determining that the majority shareholder did not have de jure control. The Court was not called upon in Duha Printers to consider the issue of how to determine whether more than one shareholder had held de jure control. This becomes clear when the Court in Duha Printers refers to the "majority shareholder" in summarizing the principles and conclusions as to control at page 838:
Summary of principles and conclusion as to control
85. It may be useful at this stage to summarize the principles of corporate and taxation law considered in this appeal, in light of their importance. They are as follows:
(1) Section 111(5) of the Income Tax Act contemplates de jure, not de facto, control.
(2) The general test for de jure control is that enunciated in Buckerfield's, supra: whether the majority shareholder enjoys "effective control" over the "affairs and fortunes" of the corporation, as manifested in "ownership of such a number of shares as carries with it the right to a majority of the votes in the election of the board of directors".
(3) To determine whether such "effective control" exists, one must consider:
(a) the corporation's governing statute;
(b) the share register of the corporation; and
(c) any specific or unique limitation on either the majority shareholder's power to control the election of the board or the board's power to manage the business and affairs of the company, as manifested in either:
(i) the constating documents of the corporation; or
(ii) any unanimous shareholder agreement.
(4) Documents other than the share register, the constating documents, and any unanimous shareholder agreement are not generally to be considered for this purpose.
(5) If there exists any such limitation as contemplated by item 3(c); the majority shareholder may nonetheless possess de jure control, unless there remains no other way for that shareholder to exercise "effective control" over the affairs and fortunes of the corporation in a manner analogous or equivalent to the Buckerfield's test. [underlining mine.]
 Thus, Duha Printers stands for the proposition that where one shareholder controls in excess of 50% of the voting shares of a corporation that shareholder will be deemed to have de jure control unless the other constating documents of the corporation, including such things as a USA, derogate from this position of control.
 However, Duha Printers was not concerned with the way in which one can determine control of a company where more than one shareholder is said to be in control. In the case of companies whose shares are widely held, an examination of the shareholders register in and of itself will not normally reveal whether any particular shareholders are in control. Therefore, I am of the view that the Duha Printers case does not assist the respondent's position.
 Most cases that address the issue of control involve situations where one or a few persons held a controlling interest. However, the critical issue here is whether a simple majority of shares held by non-residents leads to an inference of de jure control by those non-residents or whether some common connection or nexus must exist amongst those shareholders to support such an inference.
 A few cases have suggested that a common connection must exist amongst the majority shareholders in order for them to compose a "group of persons" for the purposes of the Income Tax Act. In Yardley Plastics of Canada Ltd. v.  S.C.R. 193">M.N.R. (1966), 66 DTC 5183 (Exch. Ct.), Noël J. stated that one cannot simply select an aggregation of shareholders holding more than half the voting power to be the controlling group. Justice Noël wrote at 5188:
I do not believe, as submitted by counsel for the Minister, that the latter is allowed to choose out of several possible groups any aggregation holding more than 50% of the voting power, even if the members of the group are common shareholders in both corporations and that such a group then becomes irrebuttably deemed to be the controlling group for the purposes of section 39(4) of the Act as this could lead to an absurd situation where no two large corporations in this country would be safe from being held to be associated.
 Again, in Regal Wholesale Ltd. v. The Queen (1976), 76 DTC 6146 (FCTD), Dubé J. also found that the members of a "group of persons" must have a "community of interest". Justice Dubé wrote at 6152:
[A]s defined by the Oxford and Webster dictionaries, the word "group" connotes "collective unity", "segregation from others", "having a community of interest".
 The most authoritative case dealing with control by more than one shareholder is the decision of the Supreme Court of Canada in Vina-Rug, supra. In that case, the issue was whether a company controlled by a father and his two sons was associated with a second company, more than 50% of the shares of which were held by the two sons and an unrelated party. The Court held that the second corporation was controlled by the two sons and the unrelated party because there existed between the shareholders a "sufficient common connection as to be in a position to exercise control" of the second corporation. The Court said at 196:
The learned trial judge held that John Stradwick, Jr., W.L. Stradwick and H.D. McGilvery, who collectively owned more than 50 per cent of the shares of Stradwick's Limited, had at all material times a sufficient common connection as to be in a position to exercise control over Stradwick's Limited and therefore constituted a "group of persons" within the meaning of subs. (4) of s. 39 of the Income Tax Act. I am in agreement with that finding. (underlining mine)
 Based on these cases, I agree with the appellant's submission that simple ownership of a mathematical majority of shares by a random aggregation of shareholders in a widely held corporation with some common identifying feature (e.g. place of residence) but without a common connection does not constitute de jure control as that term has been defined in the case law. I also agree with the appellant's submission that in order for more than one person to be in a position to exercise control it is necessary that there be a sufficient common connection between the individual shareholders. The common connection might include, inter alia, a voting agreement, an agreement to act in concert, or business or family relationships.
 In the present case, no evidence was adduced that would suggest the non-resident shareholders will vote as a block in the election of the directors of Alias or in other important matters relating to control of that company. The residence of shareholders alone provides no indication as to whether or not they are in agreement on the major issues relating to control of a company. The fact that there are in excess of 50% of the shareholders of Alias residing in the United States where there is no evidence that they have any common connection or indeed even know each other's identity provides no indication as to whether or not they could or would agree on any issue relating to control of the company.
 It was the respondent's position that for the purposes of the definition of CCPC in subsection 125(7), it is the residency of the shareholders that is decisive. The respondent argued that the issue of the residence of shareholders is fundamental to the determination of a CCPC, the focus being not whether the shareholders form a controlling group but rather on the residency of those holding the power to elect the board of directors.
 It seems to me that the respondent in advancing these propositions has failed to focus on the importance of the word "control" in the definition of CCPC in subsection 125(7). Residency of a cross-section of shareholders surely cannot indicate whether or not they have any power to control the corporation. Indeed, the respondent's submissions really amount to saying that if a majority of the shares of a corporation are owned by non-residents then control lies with those non-residents. The difficulty I have with this submission is that no word connoting mere ownership is used in the subsection 125(7) definition of CCPC.
 The drafter of the definition of CCPC in subsection 125(7) could have used the concept of ownership rather than the concept of control if that was intended. The words "ownership" and "owned" have been used in other parts of the Income Tax Act. For example, subsection 139A(1), which was in the pre-1972 Income Tax Act and thus predated the CCPC definition, stated:
139A. (1) For the purposes of this Act a corporation has a degree of Canadian ownership in a taxation year if throughout the sixty-day period immediately preceding that year ... not less than 25% of the issued shares of the corporation having full voting rights under all circumstances were owned by one or more individuals resident in Canada, one or more corporations controlled in Canada or a combination thereof, ...
Subsections 88(1) and (1.1) of the Income Tax Act deal with winding-up a corporation into its parent, and still provide in their opening words:
Where a taxable Canadian corporation (in this subsection referred to as the "subsidiary") has been wound up ... and not less than 90% of the issued shares of each class of the capital stock of the subsidiary were, immediately before the winding-up, owned by another taxable Canadian corporation ...
 It is worthy of note as well that by S.C. 1998, c. 19, subsections 145(2) and 145(5), subsection 125(7) was changed for taxation years commencing after 1995 by adding a paragraph to the definition of CCPC that emphasizes ownership. The original section remains unchanged as paragraph (a) of the definition, except for the addition of a reference to a new paragraph (c). In the context of the facts of this case, the change in paragraph (b) provides that all the shares held by each non-resident person shall be deemed to be held by a single non-resident individual. If that hypothetical non-resident individual would control the corporation then the corporation is not a CCPC. The wording of the CCPC definition provided by S.C. 1998, c. 19, subsection 145(2) is as follows:
"Canadian-controlled private corporation" means a private corporation that is a Canadian corporation other than a corporation
(a) controlled, directly or indirectly in any manner whatever, by one or more non-resident persons, by one or more public corporations (other than a prescribed venture capital corporation), or by any combination thereof,
(b) that would, if each share of the capital stock of a corporation that is owned by a non-resident person or a public corporation (other than a prescribed venture capital corporation) were owned by a particular person, be controlled by the particular person, or
(c) a class of the shares of the capital stock of which is listed on a prescribed stock exchange; (underlining mine)
The definition was further amended for taxation years commencing after 1999 by S.C. 2001, c. 17, subsections 113(2) and 113(4) but not in any way material to the issue dealt with here.
 The respondent argues that it is impermissible for the Court to give any consideration to subsequent changes to the definition of CCPC in subsection 125(7) and relies on subsection 45(2) of the Interpretation Act, R.S.C. 1985, c. I-21, which reads:
45. (2) The amendment of an enactment shall not be deemed to be or to involve a declaration that the law under that enactment was or was considered by Parliament or other body or person by whom the enactment was enacted to have been different from the law as it is under the enactment as amended.
 However, the Interpretation Act does not preclude the Court from drawing an inference that amendments to legislation are intended to change the legislation where the internal and external evidence warrants such a conclusion. It has been suggested that there is a presumption that changes to the wording of legislation are purposeful and that the provisions of the Interpretation Act referred to above do not preclude the Court from acknowledging that, in principle at least, the foremost purpose of amendments is to bring about a substantive change in the law. See R. Sullivan, ed., Driedger On The Construction of Statutes, 3d ed. (London: Butterworths, 1994) at page 451.
 In the present case, it seems to me that when Parliament intends that mere ownership of shares be significant in the determination of control, it uses the words "owned" or "ownership". Thus, I conclude that the word "control" in the unamended version of the definition of CCPC in subsection 125(7) did not connote mere ownership.
 I am of the view that this is a circumstance where it is obvious that a substantive change was made to the statutory provision. The change was that mere ownership of shares by a majority of non-residents would be sufficient to constitute non-resident control. This substantiates my view that the definition of CCPC in subsection 125(7) in its unamended form requires that there be a common connection among shareholders in order to show that a group of shareholders is in control.
 It should also be mentioned that the wording of the definition of CCPC in subsection 125(7) is different from other provisions of the Income Tax Act that refer to control of a corporation. For instance, subsection 39(4), the provision considered in Buckerfield's, supra, Dworkin Furs, supra, and Vina-Rug, supra, states:
For the purpose of this section, one corporation is associated with another in a taxation year, if, at any time in the year, ...
(b) both of the corporations were controlled by the same person or group of persons. (underlining mine)
 The definition of CCPC in subsection 125(7) does not mention the phrase, "group of persons" as does subsection 39(4) but only states:
"Canadian-controlled private corporation" means a private corporation that is a Canadian corporation other than a corporation controlled, directly or indirectly in any manner whatever, by one or more non-resident persons, by one or more public corporations (other than a prescribed venture capital corporation) or by any combination thereof. (underlining mine)
The respondent argued that the absence of the words "group of persons" in the definition of CCPC distinguishes it from other provisions and that cases such as Yardley Plastics, supra, and Regal Wholesale, supra do not apply (Yardley Plastics and Regal Wholesale suggest that some common connection must exist among the shareholders in order to find that de jure control exists).
 An inconsistency in the position now taken by the Minister is revealed by the view expressed by Revenue Canada at the time that the acquisition-of-control rules contained in the statute used the phrase "person or persons". Prior to 1987, subsections 111(4) and 111(5) of the Income Tax Act restricted the claiming of losses by a corporation in circumstances where "control of the corporation has been acquired by a person or persons" that did not control the corporation when the losses were incurred. In the Revenue Canada Roundtable, 1984 Conference Report (Toronto: Canadian Tax Foundation, 1985) at 816-17, the following answers were given by departmental officials responding to the following questions:
What is the Department's position concerning the acquisition of control of a loss corporation by a person or persons within the meaning of subsections 111(4) and 111(5) in the following examples?
(1) More than 50 per cent of a widely held public corporation that has losses is held by another widely held public corporation. The controlling corporation disposes of a sufficient number of shares of the loss corporation to the public so that it no longer controls the loss corporation ...
(1) If persons can be identified after the sale who own in the aggregate more than 50 per cent of the shares of the loss corporation and who also act together to control it, we would consider control to have been acquired as a result of the sale ...
The test of control is de jure control as established by the courts.
There is no jurisprudence with respect to "control ... has been acquired by a person or persons". It is our view that "persons" will be considered as having collectively acquired control where there is evidence that they have a common link or interest or they act together to control the corporation. (underlining mine)
 This is further illustrated by the Technical Notes to subsection 111(5), released when the amendment was made in 1987:
The words "person or persons" at the beginning of subsection 111(5) are changed to read "person or group of persons". This makes the terminology consistent with that used elsewhere in the Act relating to control and is not intended to cause any change in meaning.
 Of course, Technical Notes are not binding on the courts, but they are entitled to consideration. See Canada v. Ast Estate (C.A.),  F.C.J. No. 267 (C.A.), para. 27:
Administrative interpretations such as technical notes are not binding on the courts, but they are entitled to weight, and may constitute an important factor in the interpretation of statutes. Technical Notes are widely accepted by the courts as aids to statutory interpretation. The interpretive weight of technical notes is particularly great where, at the time an amendment was before it, the legislature was aware of a particular administrative interpretation of the amendment, and nonetheless enacted it.
 A similar position was put forward at the 1995 Tax Foundation Conference, Revenue Canada Roundtable, 1995 Conference Report (Toronto: Canadian Tax Foundation, 1996) at 52:10, where the position of the department was clarified as follows:
It remains our view that it is a question of fact whether persons who own the majority of voting power in a corporation constitute a group that has de jure control of the corporation. Two or more persons who become the owners of a majority of the voting shares of a corporation will generally be considered to have acquired control of the corporation where there is an agreement amongst them to vote their shares jointly, where there is evidence that they act in concert to control the corporation, or where there is evidence of their intention to act in concert to control the corporation. A group of persons would be regarded as acting in concert when the group acts with considerable interdependence in transactions involving a common purpose. A common link or interest between members of a group is required to ensure that an acquisition of control is the result of a jointly decided action, rather than a mere fortuitous event.
 Of course, statements by Revenue Canada officials are not declarative of the law. However, in the recent case ofCanadian Occidental (U.S.) Petroleum Ltd. v. The Queen,  DTC 295 (T.C.C.), Bowman A.C.J. noted that while the administrative position of Revenue Canada is not declarative of the law, it is nonetheless of assistance in circumstances where the Minister seeks to reassess the taxpayer in a manner inconsistent with its own administrative position. Associate Chief Justice Bowman wrote at 299:
The Court is not bound by departmental practice although it is not uncommon to look at it if it can be of any assistance in resolving a doubt: Nowegijick v. The Queen et al., 83 DTC 5041 at 5044. I might add as a corollary to this that departmental practice may be of assistance in resolving a doubt in favour of a taxpayer. There can be no justification for using it as a means of resolving a doubt in favour of the very department that formulated the practice.
 Quite apart from the statements of the departmental officials, I have considerable doubt that a difference exists between the meaning of the phrase "controlled by one or more ... persons" and the phrase "controlled by a person or group of persons".
 In Buckerfield's, supra at page 5303, Jackett P. said:
The word "group" in its ordinary meaning, as I understand it can refer to any number of persons from two to infinity.
In the context of control, the phrase "one or more persons" surely must mean the same thing and I am therefore of the view that the concept of "a group of persons" and the case law attendant thereon is applicable when interpreting the definition of CCPC in subsection 125(7). The significant word is "control" and in my view "control" necessitates that there be a sufficient common connection between the several persons referred to in that definition in order for there to be control by those several persons. It should be noted as well that the respondent's assertion that there is an intentional distinction in the Income Tax Act between the phrase "control by a group of persons" and the phrase "control by one or more persons" is inconsistent with published comments with Revenue Canada and the Department of Finance.
 In conclusion, it is my view that the phrase "control by one or more persons" which remains in the CCPC definition has no different meaning from the phrase "control by a person or group of persons". Hence, the case law arising from the construction of the phrase "control by a person or group of persons" is applicable. Therefore, there must be a common link or interest between members of a group, or evidence that they act together in order for control to be demonstrated. There is no such evidence here.
 The respondent further advanced an argument to the effect that the main purpose of the underlying tax advantages given to CCPCs is to encourage economic growth and increase employment in Canada. It was said that the tax advantages are restricted to CCPCs so that the tax system subsidizes the growth of primarily Canadian owned small business rather than primarily foreign owned business.
 In this connection, it should be noted that the majority of Alias's board of directors and the entire management team were residents of Canada, that Alias' principal place of business was in Canada and that product development occurred in Canada, thus suggesting that indeed economic growth and increases in employment in Canada are taking place.
 From the foregoing analysis, I am persuaded that the concept of de jure control as developed by Buckerfield's, Yardley Plastics, Vina-Rug, and Duha Printers applies fully to the definition of CCPC in subsection 125(7). In particular, de jure control includes a requirement that a sufficient common link or interest exist amongst the shareholders that compose the "group of persons", or there must be evidence that those shareholders act together to exert control over the corporation.
 In the present case, there is no such evidence. To the contrary, Alias was widely held and there is no evidence that the non-resident shareholders even knew one another. Therefore, I find that the Tax Court Judge erred in finding that the non-resident shareholders possessed de jure control of Alias.
 Having concluded that the non-resident shareholders did not possess de jure control of Alias, I will now turn to the question of whether the non-resident shareholder possessed de facto control.
De Facto Control
 The Trial Judge found it unnecessary to make any finding with respect to de facto control. Nevertheless, on appeal, the respondent advanced essentially the same argument as had been advanced at trial, and because I have concluded that the non-resident shareholders did not have de jure control, it is necessary to consider whether the non-resident shareholders had de facto control.
 The respondent in its amended reply alleged that non-residents were in de facto control of Alias because,
throughout the taxation years of Alias ending January 31, 1992 and January 31, 1993, one or more non-resident persons had direct or indirect influence that, if exercised, would result in control in fact in the years.
 The respondent argued that Silicon Graphics Inc. ("Silicon US"), a U.S. public corporation whose shares are listed on the New York Stock Exchange, was in de facto control of Alias because of a loan Silicon US made to Alias. The argument was as follows:
In December 1991, Silicon US agreed to advance up to $5,000,000 US to Alias in consideration of a security interest in all the assets of Alias and the issuance of warrants to acquire common shares of Alias. During the period in which the loan was outstanding, Silicon US determined which creditors would be paid and the amount of that payment. Alias was required to prepare a daily cash forecast to submit to Silicon US for approval. In effect, Silicon US was in control of Alias's finances. As a consequence of this indebtedness, Silicon US held de facto control of Alias at a minimum during the period in which the loan was outstanding.
 In addition, the respondent submitted that the actions and involvement of Silicon US extended beyond that necessary for the safeguarding of its rights and interests in respect of the loan. The respondent argued that other evidence pointed to the ongoing significant influence of Silicon US over Alias:
· The founder of Silicon US was a director of Alias;
· The president, chief operating officer and chief executive officer during the years under appeal was previously a senior officer of Silicon US;
· Silicon US made financial contributions to Alias for software development and marketing; and
· Alias was dependent on Silicon US given the fact that Alias software only operated on Silicon Hardware during the years under appeal;
 Because the assumption of de facto control was withdrawn by the Minister, the onus is on the respondent to establish the facts necessary to uphold this alternative basis for the assessment. See The Queen v. Bowens (1996), 96 DTC 6128 at 6129 (F.C.A.); and Pollock v. The Queen (1994), 94 DTC 6050 at 6053 (F.C.A.).
 The case law suggests that in determining whether de facto control exists it is necessary to examine external agreements (Duha Printers, supra at 825); shareholder resolutions (Société Foncière d'Investissement Inc. v. Canada,  T.C.J. No. 1568, para. 10 (T.C.C.)); and whether any party can change the board of directors or whether any shareholders' agreement gives any party the ability to influence the composition of the board of directors (International Mercantile Factors Ltd. v. The Queen (1990), 90 DTC 6390 at 6399 (F.C.T.D.), aff'd (1994), 94 DTC 6365 (F.C.A.); and Multiview Inc. v. The Queen (1997), 97 DTC 1489 at 1492-93 (T.C.C.)).
 It is therefore my view that in order for there to be a finding of de facto control, a person or group of persons must have the clear right and ability to effect a significant change in the board of directors or the powers of the board of directors or to influence in a very direct way the shareholders who would otherwise have the ability to elect the board of directors.
 The Respondent has adduced no evidence which would satisfy these criteria. There is no evidence that Silicon US as a creditor ever exercised operational control of Alias. It simply loaned money to Alias and took steps to make sure that money was only spent with a view to protecting its position as a lender. Further, the $5,000,000 bridge financing agreement lasted for only seven weeks and was repaid by the end of Alias' 1992 taxation year. Additionally, there is evidence that Silicon US did not want to be in control of Alias because it did not want to be viewed as partisan to other customers who were competitors of Alias. Silicon US never tried to install a person in management or as a director.
 The fact that the founder of Silicon US was a director of Alias out of a board of four directors is not compelling. The suggestion that the president, chief operating officer, and chief executive officer during the years under appeal was previously a senior officer of Silicon US ignores the fact that Alias itself suggested that that person assume this position. In other words, it was not something which was forced upon Alias by Silicon US.
 Further, the fact that Silicon US made financial contributions to Alias for software development and marketing and that Alias software only operated on Silicon Hardware hardly demonstrates the sort of control necessary in order to amount to de facto control.
 It would appear that the facts alleged by the respondent in relation to the loan by Silicon US simply demonstrate that Silicon US was protecting its interests as a lender to Alias. Subsection 256(6) of the Income Tax Act provides that where a party has control in fact for a period of time in order to safeguard its rights or interest, that party is deemed not to have control in fact.
 In any event, it would appear that de facto control always remained in Canada by reason of the following findings of fact made at Trial:
(e) The majority of the Board of Directors and the entire management team were residents of Canada;
(f) Alias's principal place of business was in Toronto, Ontario;
(g) The Toronto management team annually prepared a slate of people to be elected to the Board of Directors, which slate was always accepted by the shareholders.
 I conclude that Alias was a CCPC through its 1992 and 1993 years because it was not controlled directly or indirectly in any manner whatever by one or more non-resident persons.
 I therefore would allow the appeal with costs in the Tax Court of Canada and in this Court, set aside the judgment of the Tax Court of Canada dated March 28, 2001 and order that the determination of loss made pursuant to the Income Tax Act for the taxation years of Alias Research Inc. ending January 31, 1992 and January 31, 1993 be referred back to the Minister of National Revenue for reassessment on the basis that Alias Research Inc. was throughout its 1992
and 1993 taxation years a Canadian Controlled Private Corporation within the meaning of subsection 125(7) of the Income Tax Act.
"J. EDGAR SEXTON"
A.J. Stone J.A."
Marshall Rothstein J.A."
FEDERAL COURT OF CANADA
Names of Counsel and Solicitors of Record
STYLE OF CAUSE: Silicon Graphics Ltd. v. Her Majesty The Queen
DATE OF HEARING: May 13, 2002
PLACE OF HEARING: Toronto, Ontario
REASONS FOR JUDGMENT
BY: SEXTON J.A.
CONCURRED IN BY: Stone J.A.
DATED: June 17, 2002
Mr. Roger E. Taylor
Mr. Edmund C. Rowe
Mr. Paul Lefebrve For the Appellant
Mr. Harry Erlichmann
Mr. Elizabeth Chasson For the Respondent
SOLICITORS OF RECORD:
Donahue Ernest & Young LLP For the Appellant
Morris A. Rosenberg
Deputy Attorney General of Canada For the Respondent