Final Paper An Elective Entity Classification System for ...

Tax 638 Tax Research Paper

29 April 2005

Final Paper

An Elective Entity Classification System for Canada

Francis FAVRE

Table of Contents

Introduction 1

A Brief History of Pass-Through Taxation of Corporations in Canada 2

Family Corporations 2

Group Consolidation 3

Corporate Loss Transfer System 5

The United States Models 6

Elective Entity Classification – The “Check-the-Box” Rules 7

Integration for Small Business – The S Corporation 10

Conceptual Basis for Business Entity Taxation Regimes 13

Follow the Private Law 13

Public versus Non-Public Entities 14

Size of the Entity 15

Evaluation of Alternatives 16

Alternatives to Elective Entity Classification 17

Universal Entity Level Tax 17

Statutory Classification Rules 17

Policy Evaluation of Elective Entity Classification 18

Efficiency and Neutrality 18

Horizontal Equity 20

Simplicity and Certainty 20

Government Revenues 21

i

Design of Elective Entity Classification System 22

Selection of Pass-Through System 22

Scope of Eligible Entities 22

Treatment of Single Member Entities 24

Income and Loss Allocation 24

Default Rules and Mechanics of Election 25

Treatment of Conversions 26

Anti-Avoidance Rules 27

Loss Trading 27

Income Splitting 27

Implications of Elective Entity Classification on the Foreign Affiliate Rules 27

Simplification of Foreign Entity Classification 27

Increased Tax Planning Opportunities 28

Avoidance of Foreign Accrual Property Income 28

Surplus Planning 30

Federal-Provincial Tax Issues 31

Specific Provincial Concerns 31

Comparison of Loss Transfer and Elective Classification System 32

Required Changes to the Income Allocation Rules 34

Alternative Options to Address Provincial Revenue Concerns 35

Provincial Non-Conformity 35

Inter-Provincial Compensation Arrangements 36

Surtaxes and Alternative Taxes on Electing Entities 36

Conclusion 38

Bibliography 40

ii

1

An Elective Entity Classification System for Canada

Introduction

The current Canadian system for taxing business entities is governed by the private law

classification of these entities. The Income Tax Act1 establishes separate regimes for the taxation of

corporations, partnerships (or rather their members), and trusts. The Act, however, does not contain

a meaningful definition of these different entities, nor does it provide rules for determining the

classification of a particular form of carrying on business, relying instead on the private law to

establish the character of the entity. This reflects the Canadian judicial and legislative approach to

respect the non-tax law characterisation of relationships between taxpayers. This paper takes the

position that the current system for taxing business entities is flawed, since the differences in private

law treatment often do not reflect the economic substance of the entity or arrangement. Instead,

Canada should adopt an elective entity classification system that would allow eligible entities to

elect to be taxed as either a partnership or a corporation for federal (and possibly provincial) tax

purposes. An elective classification system would improve the neutrality of the Canadian tax system

by removing tax considerations from the choice of the particular form in which business is carried

on. Elective entity classification would also address two perennial issues that are often discussed in

the realm of business taxation – the desire to provide for some mechanism of loss sharing or

consolidation within corporate groups and the integration of corporate and shareholder taxation in

the small business context.

This paper will begin by providing a brief historical overview of actual and proposed rules

providing for conduit treatment of corporations, consolidation, or loss sharing. Next, the paper will

discuss the two elective classification systems provided under US tax law, the “check-the-box”

rules and the S Corporation rules. After a discussion of the policy considerations underlying the

design of a consistent tax system for business entities, the paper will evaluate the proposed elective

classification rules from a policy perspective. The balance of the paper will cover the technical

aspects of designing a workable elective entity classification system, including a consideration of

potential provincial tax concerns and possible solutions thereto.

1 R.S.C. 1985, c. 1 (5th supplement), as amended (hereinafter “the Act”).

2

A Brief History of Pass-Through Taxation of Corporations in Canada

While entity-level taxation of corporations has been the norm in Canada, the Income Tax Act

and its predecessors did at one time or another contain various provisions allowing corporations to

be treated as pass-through entities, including a basic form of elective entity classification and a

consolidation system, which are briefly summarised in this section.2

Family Corporations

The “family corporation” regime, which was in effect from 1926 to 1932, is the only true

elective classification system in Canadian history. It allowed certain active small business

corporations to elect to be treated as partnerships for tax purposes. A “family corporation” had to

carry on an active business,3 and either (i) 75% of its shares were owned by members of one family,

of whom at least one was active in the business or (ii) 80% of its shares were owned by persons

actively employed in the business or their families.4 Thus, in spite of its name, the concept covered

not only closely held family businesses, but also “incorporated partnerships” of arm’s length

persons carrying on business together. The shareholders could elect annually that, instead of the

income being taxed at the corporate level, each shareholder resident in Canada would be deemed to

be a partner in a partnership and taxed on his share of the corporation’s income according to his

interest as a shareholder. Any income allocated to a non-resident shareholder, however, remained

taxable at the corporate level.5 Dividends paid by a family corporation were taxable only to the

extent they exceeded the income that had previously been taxed in the hands of the shareholders.6

This rule implies that the corporation would have had to keep track of the income taxed to the

shareholders in a separate “surplus pool” from which dividends could then be distributed on a tax-

free basis. However, the Act did not specifically provide for such an account, nor how it would have

to be computed. The need to maintain separate surplus accounts arises because the election was an

annual one and did not trigger any adjustments on conversion from partnership status to corporate

2 This discussion will be restricted to rules of general application and will ignore the treatment of specialised entities with a restricted field of activities, such as mortgage investment corporations, mutual fund corporations, or the (former) non-resident owned investment corporations. The reason is that these entities are usually explicitly prohibited from carrying on an active business and hence do not properly fall within the purview of “business taxation”. 3 Specifically, investment income could not exceed 25% of total revenue. 4 Income War Tax Act, R.S.C. 1927, c. 97 (“IWTA”), paragraph 2(1)(d). 5 IWTA, subsection 22(1). 6 IWTA, subsection 22(3).

3

status or vice versa. This concept will be encountered again in the discussion of the S Corporation

rules below. 7

Otherwise, the provision was very straightforward and the annual election provided

shareholders with considerable flexibility in deciding which tax treatment was optimal since the

election was due thirty days after the due date for the corporate tax return in respect of the year. One

major drawback, though, was that the flow-through of losses to shareholders was not permitted.8

Thus, unlike a partnership, the family corporation was not a true conduit. Maybe for this reason

(and also because personal tax rates at that time were substantially higher than corporate tax rates),

the family corporation does not appear to have been very popular.9 In any event, the regime only

lasted for a brief period of time. The ability to elect partnership status was eliminated as of 1933,

effectively ending the regime.10 The short duration of the family corporation regime explains in part

why the rules lack in sophistication and do not address all the aspects of the tax treatment of such an

entity.11

Group Consolidation

The option to file a consolidated tax return for certain corporate groups existed from 1932 to

1952. A corporation resident in Canada could elect to consolidate all of its 100% owned

subsidiaries that were also resident in Canada, carried on the “same general class of business” as the

parent, and had the same taxation year.12 Consolidation would apply automatically to any qualifying

subsidiaries that entered the group after the election was filed. The election could be revoked at any

time, subject to the restriction that a revoked election could not be renewed for five years.13 The

effect of the election was that the parent and its subsidiaries were deemed to be a single corporation

7 On the other hand, the cost base adjustments to reflect income or losses flowed through that are a key feature of the taxation of transparent entities today were not relevant in the days of the family corporation since capital gains were not taxable. 8 This is because subsection 22(1) specifically referred to the “income of the corporation”, and the term “income” implies a positive amount. 9 There is only a single reported case dealing with the family corporation provisions: Omer H. Patrick v. MNR, 1 DTC 303 (Ex.Ct., 1935). The case dealt with whether the minister could use a spousal attribution rule that was part of the partnership provisions to attribute income allocated to the wife’s shares to her husband (a similar, although broader, rule can be found in subsection 103(1.1) of the current Act). The court held that IWTA section 22 provided a complete code for determining the income allocation, which had to be done pro rata based on share ownership. 10 The tax-free status of dividends from previously taxed income was eliminated in 1943, and the entire concept of the family corporation disappeared with the Income Tax Act of 1948. 11 Another reason is that subsection 22(4) provided the Minister with broad discretion as to the application of the family corporation rules. 12 Subsection 75(1) of the Income Tax Act of 1948. 13 Ibid., subsection 75(3).

4

that owned all the properties of the group corporations. Consolidated taxable income (or loss) was

computed as the aggregate of the taxable income and losses of the individual group companies.14

There were no special rules for computing consolidated income; however, it appears that, in spite of

the “single corporation” fiction, each individual corporation would compute its income as a separate

entity and the results would then simply be aggregated without any adjustment for intragroup

transactions. Individual group corporations could not carry over any losses, but a consolidated loss

at the group level could be carried back one year or forward three years and deducted against

consolidated taxable income. Finally, corporations electing to file a consolidated return had to pay

an additional 2% tax for the privilege of doing so.15

This “consolidation surtax” appears to have been one reason for the five year moratorium on

consolidation after a revocation of the election – otherwise, a group would have had the incentive to

file consolidated returns only during periods of losses and switch back to individual filing once all

subsidiaries were again profitable to avoid the surtax. In addition, unlike the family corporation

election described earlier, the consolidation election was only effective for any complete taxation

years beginning after the election was filed – hence, a group could not wait until the end of a

taxation year to determine whether consolidation was optimal. The requirement for 100%

ownership and concurrent fiscal years was presumably introduced for simplicity to avoid having to

deal with minority interests or stub period income. Finally, the restriction requiring all group

companies to carry on the same general class of business was part of a more general scheme of the

tax law that restricted the deductibility of losses from one line of business against the income of

another.16 More specifically, loss carryovers could only be deducted against income from the

business in which the loss was sustained.17 Applying these streaming rules at the level of a

consolidated group would necessarily have created increased complexity, and so consolidation was

restricted to corporations in the same line of business, which rendered the sourcing of losses

unnecessary.

14 Ibid., subsections 75(6) and 75(9). 15 Stephen R. Richardson, “Transfer of Deductions, Credits, or Losses within Corporate Groups: A Department of Finance Perspective,” in Report of Proceedings of the Thirty-Sixth Tax Conference, 1984 Conference Report (Toronto: Canadian Tax Foundation, 1985),at 738. 16 Ibid., at page 740. Section 13 of the Income Tax Act prevented a taxpayer from deducting a loss from another source against the income from his principal business or occupation. In addition, it was not clear whether the general definition of income permitted the netting of incomes against losses from a different source. 17 Section 26 of the Income Tax Act of 1948. A similar streaming rule applies today only in respect of loss carryforwards after a change of control – see subsection 111(5) of the present Act.

5

The consolidation election was repealed in 1952. The official explanation was that the

introduction of more generous loss carryover rules in the preceding years made consolidation

unnecessary. However, it appears that the removal of the restrictions preventing a taxpayer from

offsetting income and losses from different sources in the same year may have been more important

in reducing the perceived need for consolidation.18 It does not appear that the removal of the

consolidation option generated much discussion at the time, suggesting it was little used or did not

have a well-organised constituency.19

Corporate Loss Transfer System

In 1985, the Department of Finance released draft legislation proposing to introduce a

system of loss transfers within an affiliated group of corporations, modelled in part on the group

relief system of the United Kingdom.20 Unlike a full consolidation system, a loss transfer system

involves the elective transfer of losses among group corporations, but otherwise does not affect the

separate status of each entity. The system would have applied to a parent corporation and its 95%

owned direct or indirect subsidiaries (all of which had to be taxable Canadian corporations).21 In

determining whether the ownership requirement was met, certain types of fixed value preferred

shares were ignored. Eligible losses would be transferred on a 100% basis, i.e., without adjustment

for the actual degree of common ownership, which was one of the reasons for the high ownership

threshold.22 Transferable losses were restricted to non-capital losses for the current year; pre-

existing loss carryover balances could not be transferred. In addition, the timing rules essentially

prevented the transfer of a loss incurred in the year of acquisition.23 Both of these restrictions were

intended to prevent the “purchasing” of pre-existing losses.

The mechanism for the loss transfer was a simple joint election between the transferor (the

loss company) and the transferee. Compensating payments for the use of losses were allowed, but

18 Supra, note 15, at 741. 19 Richard J. Horwich, A Comparative Study of Consolidated Returns and Other Approaches to the Multiple Corporations Problem, Tax Law Review (New York University School of Law), vol. 20 (1964-65), at 539. 20 Canada, Department of Finance, Budget Papers, A Corporate Loss Transfer System for Canada, May 23, 1985. 21 Eligibility was measured separately at each tier, instead of considering the overall economic ownership of the parent. Thus, multi-tier structures would have qualified even if the total interest of the top corporation in the lower tier subsidiaries was less than 95%. See Stephen R. Richardson, “A Corporate Loss Transfer System for Canada: Analysis of Proposals,” in Report of Proceedings of the Thirty-Seventh Tax Conference, 1985 Conference Report (Toronto: Canadian Tax Foundation, 1986), at 12:4. 22 Ibid. 23 Ibid., at 12:9.

6

not required.24 To keep the system simple, the proposal did not provide for adjustments to the cost

base of the shares of the transferor corporation. Instead, an adjustment would be made to decrease

the loss (or increase the gain) on a disposition of shares of a non-arm’s length corporation where it

could “reasonably be considered” that a loss transfer had resulted in an increased loss or reduced

gain on the disposition.25 The rules were obviously designed for simplicity, and would have

provided considerable flexibility in managing the income and losses of the group subsidiaries since

there was no requirement for consistency from year to year. In particular, the system would have

provided significant opportunities for managing provincial tax liabilities by shifting losses from

low-tax provinces to higher tax provinces.26 The potential impact of the loss transfer system on

provincial tax planning will be explored in further detail below. Provincial concerns were a major

reason why the proposed system was never implemented.

Thus, the Canadian tax system has experimented with a variety of options for the tax

treatment of corporations, ranging from almost full conduit treatment to group consolidation and the

elective transfer of losses. Most of these alternatives never went beyond the early stages and would

not work in their original form in today’s complex tax and business environment. The reasons for

the abandonment of the family corporation rules and the consolidation election are far from clear,

although neither regime appears to have been a central feature of the tax system of its time. The

corporate loss transfer system was not implemented due in large part to provincial revenue

concerns. Thus, we have to look abroad, and specifically to the United States, to analyse a mature

system of elective classification of business entities.

The United States Models

The US tax system currently provides three alternative models to deal with entity taxation:

the “check-the-box” entity classification system, which allows a wide variety of non-public entities

to elect their tax treatment as either a corporation or a partnership, the S Corporation rules, which

allow qualifying domestic private corporations to elect to be taxed as pass-through entities, as well

as a sophisticated consolidation system for corporate groups. This section will provide an overview

of the first two systems.27

24 Ibid., at 12:11. 25 Ibid., at 12:12. 26 Ibid., at 12:14. 27 A review of the US consolidated return rules is beyond the scope of this paper.

7

Elective Entity Classification – The “Check-the-Box” Rules

The “check-the-box”28 regulations were introduced in 1997 to simplify the classification of

unincorporated business entities as either corporations or partnerships for US federal tax purposes.

Prior to these regulations, entity classification was based on a “corporate resemblance” test which

evaluated entities based on criteria designed to capture the “substance” of the corporate form. This

test had become unwieldy to apply and had degenerated into a de facto elective system. The elective

classification rules are not a separate statutory regime within the Internal Revenue Code,29 but are

merely administrative regulations interpreting the definitions of “corporation” and “partnership” for

federal tax purposes.30 While it may seem surprising that such a seemingly fundamental change as

the introduction of an elective classification system could be introduced without clear statutory

authority, and there was initially some discussion as to whether the regulations were ultra vires, the

validity of the rules was never seriously challenged.

The first step in determining whether the elective classification rules apply is to establish

whether a separate entity exists under US federal tax law.31 This determination does not depend on

whether the organisation is recognised as an entity under local law,32 and the guidance provided in

the regulations is limited, although certain contractual arrangements, such as joint ventures, could

create a separate entity if a certain level of activity is reached.33 If a separate entity exists, it will be

a “business entity” unless it is properly classified as a trust34 or otherwise subject to special tax

treatment. Certain entities are deemed to be corporations and are not eligible to elect their tax status.

These “per se corporations” include entities incorporated under federal or state law, joint stock

companies, banks and insurance companies, as well as a list of foreign entities that are considered

sufficiently similar to a US corporation to be automatically classified as such. The per se

corporations list was included in the regulations because allowing entities that are prima facie

28 The moniker “check-the-box” arose because the classification election is made by checking the appropriate box on the election form (Form 8832). 29 Internal Revenue Code of 1986, as amended (hereinafter “IRC”). 30 IRC s. 7701(2) and (3). Section 7701 is a general definition section roughly analogous to subsection 248(1) of the Act. 31 Allan R. Lanthier and Kerry L. Plutte, “International Hybrids: Pitfalls and Practice,” in Report of Proceedings of the Fifty-First Tax Conference, 1999 Conference Report (Toronto: Canadian Tax Foundation, 2000), at 46:10. 32 Treasury Regulation 301.7701-1(a)(1). 33 Treas. Reg. 301.7701-1(a)(2). The tests in this regulation are similar to those used in establishing whether a partnership exists under Canadian law. 34 Note that “commercial trusts” (i.e. trusts used to carry on a business) will be classified as business entities eligible for the check-the-box rules, and will not be taxed as ordinary trusts. Treas. Reg. 301.7701-4(b).

8

corporations to be taxed as partnerships would have exceeded the authority of the regulations,

which are supposed to merely interpret the IRC and cannot change the law.35

Any business entity that is not a per se corporation (an “eligible entity”) is allowed to elect

its tax classification. An eligible entity with at least two members can elect to be taxed as either a

corporation or a partnership, while an entity with a single owner can elect to be classified as a

corporation or to be disregarded as an entity separate from its owner.36 In the latter case, its

activities will be treated in the same manner as a sole proprietorship, branch, or division of the

owner. If the entity does not file an election, the default rules will apply. The default treatment for a

domestic entity is (i) a partnership if it has two or more members or (ii) disregarded if it has a single

owner. For a foreign entity, the default treatment is (i) a partnership if it has several members and at

least one member has unlimited liability, (ii) a corporation if all members have limited liability, or

(iii) disregarded if it has a single owner with unlimited liability.37 If an entity wants to achieve a

status other than the default, it must file an affirmative election to change its status, which will be

effective on a prospective basis.38 Once an affirmative election has been made, it cannot be changed

for five years after the effective date unless there has been a change of control and the Internal

Revenue Service (“IRS”) agrees to permit a new election.39

The regulations also specify the tax consequences of a change in status. If an entity

classified as a partnership elects to be taxed as a corporation, the “partnership” is deemed to

contribute all of its assets and liabilities to a corporation in exchange for shares, and immediately

thereafter, the partnership is deemed to liquidate and distribute the shares of the notional

corporation to its members. If an entity classified as a corporation elects to be treated as a

partnership, there is a deemed winding-up and liquidation of the corporation followed by a

contribution of the distributed assets by the former shareholders to a newly formed partnership.

Analogous rules apply to the conversion from a corporation to a disregarded entity and vice versa.

The regulations specifically provide that these deemed transactions will be subject to all the tax

35 The check-the-box rules purport to interpret the term “association” used in the legislation. IRC s. 7701(3) states that the term ‘corporation’ includes associations, which are thus taxable as a corporation. However, a state law corporation falls within the plain meaning of the term “corporation” and hence there is no interpretive leeway to treat it as anything else but a ‘corporation’ for tax purposes. 36 Treas. Reg. 301.7701-3(a). 37 Treas. Reg. 301.7701-3(b). 38 The effective date of the election cannot be more than 75 days prior to the date of filing. 39 Treas. Reg. 301.7701-3(c).

9

consequences that would apply had they actually occurred.40 Thus, a conversion will usually be a

taxable event.

The “check-the-box” rules do not create a separate tax regime; the existing rules for

corporations or “ordinary” partnerships will apply to the electing entities based on their choice or

default classification. 41 Accordingly, the application of the rules is fairly straightforward, and the

most important consideration will usually be the tax consequences, if any, of the deemed

conversion. The rules are also very flexible – except for the five year lock-in after an affirmative

election has been made, there are few restrictions – no minimum ownership or residency

requirements or restrictions on the types of owners apply.

Nevertheless, there is one important exception to this flexibility: all publicly traded entities

will be taxed as corporations, even if they elect to be treated as a partnership. The IRC specifically

provides that any publicly traded partnership will be taxed as a corporation.42 A partnership will be

considered publicly traded if its units are traded on an established securities market or are readily

tradable on a secondary market. An exception to this rule applies if at least 90% of the gross income

of the partnership is derived from property income (interest, dividends, rents), real estate activities,

or resource activities. Thus, the check-the-box regime effectively applies only to private entities or

the subsidiaries of public entities. As a result, a phenomenon akin to the proliferation of “business

income trusts” and similar vehicles to avoid the corporate level tax for publicly traded businesses

would be impossible in the United States.43

In spite of this restriction, the check-the-box rules provide a very flexible regime to achieve

conduit treatment for a wide variety of entities and groups in situations where neither the US

consolidated return rules nor the S Corporation rules would apply. Since per se corporations are

precluded from making a classification election, the entity of choice is the limited liability company

(“LLC”), a hybrid entity with characteristics of both corporations and partnerships, but which

provides limited liability to all its members and can generally be owned by a single member. These

40 Treas. Reg. 301.7701-3(g). 41 Subchapters C and K of the IRC, respectively. 42 IRC s. 7704. It should be noted that “partnership” would include a business trust, limited liability company, or other entity that would be taxed as a partnership in the absence of this deeming provision. 43 IRC s. 7704 was enacted in 1987 precisely as a response to the perceived threat of disincorporation resulting from the proliferation of publicly traded “master limited partnerships,” which achieved similar results to Canadian income trusts. The speed with which the US Congress acted to shut down the publicly traded partnerships is in marked contrast to the apparent relative indifference of the Canadian government to the proliferation of the income trusts. Without the safeguard of the publicly traded partnerships rule, it is inconceivable that the check-the-box rules would have been introduced since they would have allowed for the formation of publicly traded LLCs taxed as partnerships, resulting in a severe erosion of the US corporate tax base. See infra, note 59, at 542.

10

characteristics make the LLC simpler to use than the more traditional limited partnership, which

requires at least two members, one of which must have unlimited liability.

Integration for Small Business – The S Corporation

The S Corporation rules were introduced in 1958 to allow small businesses to avoid the

double taxation inherent in the US classical corporate tax system, which provides no recognition for

the corporate tax paid when corporate income is distributed to the shareholders as a dividend.44 An

eligible “small business corporation” may elect to be treated as a pass-through entity. The term

“small business” is somewhat misleading, as there are no restrictions on the size of the business per

se (as measured by assets, revenues, or taxable income). However, an S corporation must meet the

following conditions:45

(i) The corporation must be a domestic corporation;

(ii) The number of shareholders is limited to a maximum of 75;46

(iii) The shareholders are restricted to individuals, certain qualifying trusts, and certain

tax exempt organisations;

(iv) Non-resident shareholders are not permitted; and

(v) The corporation cannot have more than 1 class of shares.47

In addition, financial institutions and certain specialised corporations are not eligible.

An S Corporation can elect to have any wholly-owned subsidiary effectively disregarded.

The wholly-owned subsidiary is not treated as a separate entity for tax purposes and all of its assets,

liabilities, income, and deductions are attributed to the parent corporation. In effect, the treatment of

wholly-owned subsidiaries is analogous to that of single-member disregarded entities under the

check-the-box rules. On the other hand, there is no mechanism to consolidate less than 100% owned

corporate subsidiaries, even if all of their shareholders are S Corporations.

An eligible corporation can file an election to be treated as an S Corporation if all

shareholders consent to the election. The election is effective for the taxation year it is made

provided it was filed within the first 2½ months of the year; otherwise, it takes effect for the

following year. The election remains in effect until it is terminated voluntarily or involuntarily. The

44 There is no equivalent to the Canadian dividend tax credit in the US tax system. 45 IRC s. 1361. 46 Husband and wife are treated as a single shareholder for the purpose of determining the number of shareholders. Thus, technically, the corporation could have up to 150 shareholders. 47 Differential voting rights are ignored for the purpose of determining whether a separate class of shares exists. Thus, an S Corporation could have voting and non-voting common shares, but it may not issue preferred shares.

11

election can be revoked at any time by a majority of the shareholders, but, once revoked, cannot be

renewed for another five years unless the IRS consents to the renewal. The election will also be

terminated involuntary if the corporation fails to meet any of the criteria listed above or has excess

passive income. Unlike a status change under the check-the-box rules, the conversion to or from S

Corporation status does not result in a taxable liquidation and hence generates no immediate tax

consequences. However, losses incurred during the period the corporation was a taxable corporation

cannot be carried forward to reduce the income of an S Corporation and vice versa.48

An S Corporation is not subject to corporate tax. Instead, the corporation’s taxable income is

computed at the entity level and then flowed through to the shareholders.49 The shareholders then

include their share of the corporation’s income or loss in their personal income pro rata to their

shareholdings. As in a true partnership, income from specific sources retains its character in the

hands of the shareholders as if it had been earned directly. Losses are deductible only to the extent

of the shareholder’s cost base in the shares and indebtedness of the corporation. Excess losses may

be carried forward indefinitely and become deductible if the shareholder’s basis increases in the

meantime (as a result of income allocations or additional investments). These rules are similar to the

at-risk rules that apply to Canadian limited partners.50 The cost base to a shareholder of the shares

of an S Corporation is adjusted in a manner similar to that of a partnership interest. Income

allocated to the shareholder increases the cost base, while deductions, losses, distributions (to the

extent not included in income), and non-deductible expenses that are not on capital account

decrease the cost base. The cost base cannot be decreased below zero (i.e. negative basis is not

possible). Any excess deductions will reduce the basis of any debt of the corporation owned by the

shareholder.51 As mentioned earlier, losses in excess of the basis of shares and debt are suspended

and must be carried forward.

The tax treatment of a dividend from an S Corporation depends on whether the corporation

has tax retained earnings from periods not covered by the election. If the corporation does not have

any pre-conversion retained earnings,52 all distributions are treated as a basis reduction, and

distributions in excess of basis create a deemed capital gain. The treatment becomes more complex

48 IRC s. 1371(b). 49 IRC s. 1366. 50 Subsection 96(2.1). Note, however, that, unlike Canada, the US has a general rule that partners in a partnership cannot deduct losses in excess of their basis in the partnership interest. This is not true in Canada for general partners, who can create a negative adjusted cost base by deducting losses in excess of their investment. 51 IRC s. 1367. 52 Termed “accumulated earnings and profits” in the legislation.

12

if the corporation has accumulated retained earnings from any period during which it was a regular

taxable corporation. Generally, dividends from surplus accumulated during the time the corporation

was an S Corporation are tax-free, but reduce the basis of the shares. Dividends in excess of this

tax-free surplus are treated as ordinary taxable dividends to the extent of any pre-conversion

retained earnings. Dividends in excess of both surplus pools are again treated as a basis recovery

and will generate a deemed capital gain if the basis in the shares would otherwise go negative. The

requirement to maintain surplus accounts for both the pre-conversion and the post-conversion

period arises because the conversion itself does not result in a taxable winding up of the

corporation. In the absence of such tracking accounts, the surplus of a taxable corporation could be

“stripped out” without personal tax consequences simply by converting the taxable corporation into

an S Corporation.

Another issue from the absence of a taxable liquidation arises when the corporation has

accrued gains on its assets at the time of conversion, which would be subject to corporate tax on an

actual winding up. If such accrued gains are realised by the S corporation within 10 years of the

conversion, a special tax at the highest corporate tax rate is imposed on these deferred gains.

Notwithstanding the general rule against carrying forward pre-conversion losses, such losses are

allowed against the income derived from the realisation of pre-conversion gains.53 Again, this is

basically an anti-avoidance rule to prevent the avoidance of the corporate tax on liquidation by first

converting the liquidating company to an S Corporation.

In summary, the S Corporation rules are designed to offer a “simplified” pass-through

regime for certain closely held corporations. Some of the restrictions, such as those on the number

of shareholders and the prohibition of multiple classes of shares, appear to be driven by

administrative convenience. Allowing the issuance of preferred shares could create problems in

determining the proper allocation of income to shareholders, as will be discussed in more detail

later. Some of the other complexities of the rules are a consequence of the desire to avoid triggering

a taxable liquidation on conversion. The requirement to keep track of pre-conversion surplus

accounts and unrealised gains would not exist if conversion were a taxable event. The basis

adjustments become necessary to reconcile gains and losses on disposition with the income

allocated to the shareholders and hence prevent the double-taxation of gains and the “double-

deduction” of losses. The basis adjustments also serve as an “at-risk rule” to deny deductions in

excess of economic investment and previously taxed income. Finally, it is interesting to note that 53 IRC s. 1374.

13

elective regimes as diverse as the former Canadian consolidation rules, the check-the-box rules, and

the S Corporation all provide for a five year moratorium on the renewal of a revoked election – the

designers appear to have had a common concern about taxpayers picking one regime in some years

and a different one in others based on which regime would provide the most tax benefits in a given

year.54

Conceptual Basis for Business Entity Taxation Regimes

Follow the Private Law

Currently, the Canadian tax system taxes different forms of carrying on business strictly in

accordance with their classification under the relevant non-tax law (common law or applicable

statutes). Once an entity has been determined to be a corporation or partnership (or trust) under the

applicable private law, the tax treatment follows automatically without questioning the substance of

the form chosen. The conduit treatment of partnerships reflects the “aggregate” view of partnerships

under common law – i.e., a partnership is not a separate legal entity but merely a relationship

between persons carrying on business in common. A corporation, on the other hand, has a separate

legal existence by statute and thus is treated as a separate taxpayer.55 This strict adherence to private

law does have some advantages, the most important of which are certainty and simplicity. Entity

classification has never been an issue for domestic purposes, and a taxpayer choosing to carry on

business in non-corporate form never had to worry that the arrangement might attract entity-level

tax. However, the existence of a partnership has been frequently litigated, usually in connection

with “tax-shelter” type arrangements involving the “acquisition” of losses.56

On the other hand, the classification of foreign entities for Canadian tax purposes can be

problematic if the legal attributes of these entities are materially different from Canadian

corporations, partnerships, or trusts.57 Since there are no statutory classification rules in the Act,

taxpayers must generally rely on the administrative positions of the Canada Revenue Agency

54 The family corporation rules are an exception in that they provided for an annual after-the-fact election and hence offered taxpayers all the flexibility they could ask for in optimising their tax status on an annual basis. However, it is likely that this was more due to a lack of experience on the part of the drafters than to a conscious policy choice. 55 Interestingly, the tax treatment of trusts does not strictly follow the common law: a trust is treated as a separate taxpayer with only limited conduit treatment even though, at common law, a trust is simply a fiduciary relationship between a trustee holding property for the benefit of other persons, and would not be considered to be a separate entity. 56 See the cases of Backman, 2001 DTC 5149 (SCC), and Spire Freezers, 2001 DTC 5158 (SCC). 57 The classification of foreign entities is relevant for the purposes of the Canadian foreign affiliate regime.

14

(“CRA”), 58 which are not always well thought out or even consistent. In case of disagreements, the

taxpayer could of course decide to litigate the issue, but this would impose additional costs and

uncertainty and is not a feasible option at the planning stage. Further, as a matter of economic

substance, there appears to be little justification in basing the tax treatment on the legal status of an

entity. Apart from liability issues, there is little to distinguish a small owner-managed corporation

from a proprietorship or general partnership. There is even less distinction between a large publicly

traded limited partnership or “income trust” and a regular public corporation. Thus, while it may be

appropriate to tax certain businesses as entities and allowing conduit treatment for others, it is not at

all obvious that this distinction should turn on the private law classification of the entity.

Public versus Non-Public Entities

One possible way to draw the line for imposing an entity level tax is the distinction between

closely held and widely held or publicly traded entities. Under such a system, the corporate tax

would be imposed on all publicly traded entities, irrespective of their classification under private

law. Non-public entities would be treated as conduits or allowed to elect their tax treatment. With

the combination of the S Corporation, the check-the-box rules, and the publicly traded partnership

rules, this is largely the de facto system currently adopted by the United States. Differential tax

treatment between public and non-public entities could be justified based on the benefits principle

of taxation, i.e., publicly traded entities receive benefits from the government that justify imposing a

higher tax burden. Publicly traded entities have the ability to raise large amounts of capital from the

general public. The investors, who ultimately bear most of the burden of the additional tax, enjoy

the protection of the securities laws, well-established rules of corporate governance, and the ability

to freely trade their interests in a liquid and regulated market, without regard to controlling

owners.59

In addition, drawing the line for entity taxation at public trading should also mitigate

efficiency concerns. Specifically, it is very unlikely that an organisation would forgo the benefits of

public trading just for the tax benefits of conduit treatment. The ability to monetise their investment

in an IPO would be expected to far outweigh any concerns about additional taxation in the owners’

decision to take a private business public. Similarly, the vast majority of public corporations will

58 See IT-343R - Meaning of the term corporation, as well as numerous technical interpretations on the status of specific foreign entities. 59 Victor E. Fleischer, “If it Looks Like a Duck: Corporate Resemblance and Check-the-Box Elective Tax Classification,” 1996 Columbia Law Review 518, at 554.

15

not forgo public trading in order to provide conduit treatment to their investors.60 Thus, public

entities are more likely to “accept” the higher tax burden imposed by entity level tax and also have

the financial capacity to bear the costs. Entity level tax could then be viewed as the price paid for

accessing the public capital markets.61 From a practical perspective, a distinction based on public

trading should be relatively straightforward to administer and difficult to plan around. While there

would be a grey area for entities that are not listed on a public stock exchange or organised dealer

network but could still be considered to be actively traded, the problem appears manageable. The

test would essentially revolve around the ability of an investor to dispose of an interest within a

short time frame, with little personal effort, without the approval of any other members, and with no

substantial illiquidity discount. Even if the rule was restricted to entities listed on a recognised

exchange (e.g., the TSX, Canadian Venture Exchange, NASDAQ, etc.), it should still be reasonably

effective as most public entities would not make do with being listed on some informal bulletin

board.

Size of the Entity

An alternative threshold for entity level taxation would rely on size, rather than public

trading. Under this option, “large” businesses would be subject to entity level taxation while small

businesses would be treated as conduits.62 One proposal suggested setting the threshold at a specific

level of gross revenue. Gross revenue was chosen instead of other measures, such as profitability or

asset-based criteria, because it is not usually subject to large fluctuations and is relatively difficult to

manipulate.63 The threshold level of gross revenue would be chosen based on a number of factors,

including the estimated revenue effects of the change and the number of entities that would have to

change their tax status under this new system.64 In addition, a transition rule would be included to

prevent one-time anomalous fluctuations in sales from causing a status change.65 The policy

rationale for basing the tax treatment on size is presumably that larger businesses have more market

power and other influence, and thus it is appropriate to tax them more heavily. Supporters of this

60 Witness the problems of certain Canadian income trusts in restricting the percentage of non-resident owners to comply with the mutual fund trust requirements. One way of discouraging tax arbitrage is to ensure that the entity must bear the full economic consequences of the form chosen. 61 Supra, note 29, at 555. 62 Curtis J. Berger, “W(h)ither Partnership Taxation?,” Tax Law Review (New York University School of Law), vol. 47 (1991), at 105. 63 Ibid., at 164. 64 Ibid., at 165. 65 Ibid., at 169.

16

proposal reject public trading as an appropriate threshold because doing so would exclude otherwise

highly profitable closely held entities from the entity level tax. In addition, linking entity level tax to

public trading might discourage certain businesses from seeking additional investment in the public

market if the cost was the loss of conduit treatment.66 While it is unlikely that an established public

corporation would be taken private merely to achieve flow-through treatment, the loss of the tax

benefits associated with conduit treatment could act as a disincentive at the margin in the decision

to take a private firm public. Such a result would reduce the efficiency of capital markets and would

result in a higher cost of capital for the affected businesses.

Evaluation of Alternatives

Of all the potential bases for entity level taxation, public trading is the soundest both from a

policy perspective and for practical reasons. From a policy perspective, the ability to access public

capital markets and to easily monetise the investment of the business’ founders represents a

substantial benefit that it may be appropriate to tax. Basing the corporate tax on public trading

would also create fewer distortions than the alternatives. The current system of following the legal

form of the entity encourages businesses to convert to alternative forms, such as trusts or limited

partnerships, and poses a threat to the corporate tax base. While size may appear at first to be an

appealing basis for imposing an additional tax, the theoretical justification for the distinction is not

well founded, as size itself does not create any special benefits. Although the owners of a large

business tend to be wealthier than those of a small one, the issue really is one about the

progressivity of the individual tax system, rather than a basis for imposing an entity-level tax.

Further, while the distinction between a closely held and a publicly traded entity creates a

meaningful discrete cut-off point, any threshold based on size will necessarily be arbitrary and will

not reflect a fundamental underlying difference. Lastly, as mentioned earlier, a distinction based on

public trading should be easier to enforce than one based on size and is inherently simpler to apply.

It should also create fewer distortions in economic decision making, as the demand for public

trading would seem to be relatively inelastic.67

66 Ibid., at 162. 67 For example, if entity level taxation were based on size, there would be an incentive for a large partnership, e.g., a national law or public accounting firm, to split into a group of local partnerships (e.g. one in each city) with some kind of loose affiliation. Such splits could not be prevented by affiliation rules (such as the associated corporations rules) since the partners would not be related. There are presumably good business reasons for having national partnerships (such as spreading the risk of unlimited liability among more partners), and hence creating an incentive for the fractionalisation of large partnerships is economically undesirable.

17

Alternatives to Elective Entity Classification

Universal Entity Level Tax

Under this option, all business entities would be taxed as corporations regardless of their

legal nature or ownership characteristics. Such a system would provide simplicity and would protect

the corporate revenue base against the threat of “disincorporation” evident in the current popularity

of income trust arrangements. However, in the absence of a relieving mechanism, such as a

corporate loss consolidation regime, entity taxation would worsen underintegration and might

encourage carrying on business as sole proprietorships or unincorporated divisions, even if other

arrangements would be more suitable from a commercial perspective. It would also encourage the

avoidance of partnership status in favour of less well defined arrangements, such as joint ventures.

These incentives would necessitate potentially arbitrary cutoff rules to determine when the

threshold for entity taxation has been passed. Accordingly, universal entity level tax is not a sound

basis for business taxation.

Statutory Classification Rules

Statutory definitions of “partnership” and “corporation” would be introduced based on

criteria designed to capture the difference in economic substance between these two forms of

carrying on business. Such rules might reduce planning opportunities involving the use of different

legal forms where the economic substance of the arrangement is the same. They could also clarify

the basis for the characterisation of foreign entities, which is currently based on CRA administrative

positions and court decisions. However, such a system is also likely to lead to increased complexity

and uncertainty, as case-by-case evaluations might be necessary. Attempts to create bright line tests

would necessarily result in a form-driven set of rules encouraging tax planners to structure entities

to fit within the desired definition. The US experience in this domain suggests that a statutory

classification scheme is likely to be unworkable and futile.68

68 Former Treasury Regulation 301.7701-2(a) provided a four factor test to evaluate corporate status (known as the Kintner regulations). An entity would be classified as a corporation if it had at least three of the following characteristics: continuity of life, centralised management, limited liability, and free transferability of interests. The evaluation of these factors was very form driven without any evaluation of the importance of each characteristic and encouraged taxpayers to draft the organisational documents of the entity to achieve their preferred tax status. As a result, by the time it was abandoned in 1996, the system had become a de facto elective system. See, infra, note 72, at 125.

18

Policy Evaluation of Elective Entity Classification

Efficiency and Neutrality

An elective classification system would improve the neutrality of the tax system with respect

to the different ways of structuring a business organisation. Within a corporate group, elective entity

classification would be a straightforward way of achieving the consolidation of income and losses at

the level of the economic, rather than the legal, entity. Removing tax considerations from the factors

driving the choice of business structure will leave corporate groups free to choose the structure that

best satisfies their business objectives. There will no longer be an incentive to structure transactions

via various joint venture or cost sharing agreements which are chosen to achieve effective

consolidation, but which may not be the optimal structure from a business point of view. In

addition, the various artificial transactions currently used to achieve effective consolidation of

losses will no longer be required, freeing up resources for more productive activities. Similarly, tax-

driven corporate restructuring, such as amalgamations and wind-ups, will be reduced. This may be

beneficial to the extent the retention of separate entities is more efficient from a business

perspective, such as to isolate risky activities in separate subsidiaries to shield the main business

from the liabilities of a particular business venture. Finally, the effective consolidation provided by

an elective classification system would facilitate the formation of separate corporations to carry on

particular divisional responsibilities. In addition to limited liability, the following advantages of

using separate subsidiaries instead of divisions have been cited:69

1. The ability to provide for divisional management and employee share participation, as well

as greater flexibility in employee identification and remuneration packages. These benefits

arise because stock based compensation is more effective if the performance of the

corporation is closely tied to employee effort. Generally, this is more likely to be the case

where the entity is fairly small and operates in a single line of business.

2. The ability to conclude separate labour agreements at the entity level.

3. Increased opportunities to obtain separate project financing because the assets and liabilities

of the particular project are legally segregated from those of other projects.

69 Glen E. Cronkwright, “The Utilization of Losses in Corporate Groups and Further Relief that Might be Taken,” in Report of Proceedings of the Thirty-First Tax Conference, 1979 Conference Report (Toronto: Canadian Tax Foundation, 1980), at 324.

19

Thus, elective tax classification would increase the flexibility of corporate groups to structure their

affairs in the most efficient manner without being unduly concerned with the tax treatment of a

particular structure.

A similar argument can be made at the level of the sole proprietor or small business. While

the Canadian tax system provides many benefits to incorporated small businesses,70 it is fair to say

that most of these benefits only apply to profitable businesses.71 Thus, there is still a tax incentive to

start a new venture as a proprietorship or partnership to personally access the losses during the start-

up phase. However, given the high risk of bankruptcy facing most new businesses, these legal forms

characterised by their unlimited liability are likely not optimal from a non-tax perspective. An

elective classification system would allow new businesses to be carried on under the limited

liability protection of the corporate form without giving up the ability to offset the start-up losses

against the owner’s personal income. Thus, an elective classification system could encourage risk

taking and the starting of new businesses, which would be beneficial to the economy.

Nevertheless, elective tax classification has also been criticised on efficiency grounds.

Making an election may create additional complexities for taxpayers, and there are additional costs

in determining which status would be optimal for a particular taxpayer. In addition, there is a risk

that ill-advised taxpayers may make the wrong election and could be stuck in a particular tax regime

because of the tax costs that a conversion at a later date may impose.72 However, it is a general

characteristic of a complex tax system that sophisticated taxpayers will derive more benefits from

the system than unsophisticated ones. Similarly, the choice of business form is already a complex

decision that requires evaluating many tax and non-tax factors, so that the ability to make a

classification election should not add significant additional complexity.

70 The tax deferral for active business income below $300,000 retained in a corporation and the $500,000 lifetime capital gains exemption being the major ones. 71 Refundable tax credits for R&D are one major tax benefit that also applies to corporations in a loss position. The other is the ability to deduct a capital loss on the disposition of debt or shares of a small business corporation against ordinary income as an “allowable business investment loss.” 72 George K. Yin, “The Taxation of Private Business Enterprises: Some Policy Questions Stimulated by the “Check-the-Box” Regulations,” Southern Methodist University Law Review vol. 51, at 130.

20

Horizontal Equity

Elective tax classification would improve horizontal equity between corporate groups that

can already effectively consolidate their income and losses by operating as a single entity or by

utilising various flow-through vehicles and those that cannot do so, be it because their industry is

particularly exposed to liability issues (e.g. natural resources, real estate) or because of regulatory

restrictions (e.g. banking and insurance). In addition, as a matter of basic corporate tax theory, the

separate legal existence of many corporations in a group should not generally affect the tax liability

borne indirectly by the ultimate investors.73 If the corporate tax is viewed as essentially an advance

withholding of investor-level tax, the number and characteristics of the layers of entities between

the operations and the ultimate individual investors should not affect the amount of tax withheld.

Similarly, if the corporate tax is viewed as a tax on excess profits or economic rents, the proper tax

liability should be determined based on the income of the economic, rather than legal, entity.74

Thus, elective tax classification, like a formal consolidation regime, would improve horizontal

equity between different forms of carrying on business that are economically equivalent.

Simplicity and Certainty

The case for elective classification as enhancing simplicity is not nearly as strong in Canada

as it was in the United States. In the US, the move to the check-the-box rules was justified on the

grounds of improving simplicity (and to a lesser extent certainty) in an entity classification system

that had degenerated into a set of form-driven rules where both taxpayers and the revenue

authorities spent significant time designing and analysing the characteristics of legal entities to

achieve the desired tax treatment. Canada does not face a similar problem, at least at the domestic

level: entity classification strictly follows the non-tax law and is thus both simple and certain – if an

entity is a corporation, partnership, or trust under the relevant federal or provincial law, it will

almost always be taxed as such.75 In the realm of the foreign affiliate rules, on the other hand, entity

classification can be a real problem, as the jurisprudence is limited and the administrative positions

of the CRA do not provide a very clear framework either. In this area, an elective classification

system would have significant benefits in terms of simplicity and certainty. Simplicity may also be

73 Robert Couzin, “Income Taxation of Groups of Corporations: The Case for Consolidation,” in Report of Proceedings of the Thirty-Sixth Tax Conference, 1984 Conference Report (Toronto: Canadian Tax Foundation, 1985), at 730. 74 Ibid., at 727. 75 This statement will always be true for corporations and trusts. On the other hand, the tax status of a partnership is sometimes challenged, but usually only in cases where a blatant tax avoidance motive exists (e.g. the purchasing or transfer of pregnant losses).

21

enhanced indirectly to the extent elective classification removes the need for complex structures and

transactions designed to achieve effective consolidation of income and losses within a corporate

group.

Government Revenues

By its very nature, any system that allows taxpayers to elect a particular tax treatment will

always result in lower government revenues relative to the case were the tax treatment is imposed

externally. Taxpayers will not choose to increase their tax liability, unless they are ill-advised or

lack complete information about the consequences of their choice.76 The question then becomes

essentially one of the magnitude of the expected revenue loss, and whether the efficiency and equity

benefits justify the cost in foregone government revenues. It has often been noted in discussions

about the need for a group relief system that most Canadian corporate groups are able to effectively

utilise their losses by relying on ‘in-house” consolidation mechanisms explicitly or implicitly

sanctioned by the CRA.77 If this is correct, the additional revenue loss from elective classification

may not be very large. On the other hand, under the current system, in-house loss transfers are only

allowed within an affiliated group, and, since most taxpayers will request a ruling before proceeding

with a given scheme, the CRA has a relatively tight control over the level of de facto loss

consolidation. The proposed elective classification system would apply to all non-public entities and

would not be restricted to a group with high levels of common ownership, as were previous loss

transfer proposals. Thus, the revenue cost will both be greater and more difficult to estimate

accurately. In addition, the ability of electing private businesses to flow out losses to individuals

would result in additional revenue costs. Nevertheless, the revenue cost should be manageable

provided accumulated losses cannot be flowed out by means of an election. At the individual level,

loss limitation rules such as the at risk rules and proposed reasonable expectation of profit

requirements78 should provide appropriate safeguards. In the end, the decision to adopt an elective

classification system will require a balancing of competing factors, with the potential revenue loss

being just one of them.

76 One way of preventing taxpayers from making their choice in full knowledge of the tax consequences is to require that any election be made ex ante instead of after the fact, and by restricting the flexibility to change or revoke the election. These considerations will be discussed in more detail in the design section. 77 See, for example, footnote 5 in Nicholas LePan, “Federal and Provincial Issues in the Corporate Loss Transfer Proposal,” in Report of Proceedings of the Thirty-Seventh Tax Conference, 1985 Conference Report (Toronto: Canadian Tax Foundation, 1986), at 13:4. 78 Proposed section 3.1 of the Act.

22

Design of Elective Entity Classification System

Selection of Pass-Through System

The first issue in designing an elective classification system is whether to create a separate

pass-through regime for electing entities or whether the existing partnership provisions of the Act

should apply.79 Using the existing partnership provisions has several advantages over designing a

new system. First, introducing two different sets of rules to achieve the same goal, i.e. a conduit

system for the allocation of income and losses to members of an entity, is inherently illogical and

potentially confusing to taxpayers. Second, a dual system would not achieve the goal of neutrality,

as one would expect one system to be more advantageous in a particular situation than the other.

Thus, taxpayers would still have the incentive to structure their affairs with the goal of falling into

one system or the other. This defeats the purpose of elective classification, namely the unlinking of

entity choice from tax considerations. Finally, using the existing rules enhances the certainty of

application, as taxpayers can rely on existing jurisprudence and administrative interpretations.

Scope of Eligible Entities

The next step is to determine which entities should be able to make the election. The US

check-the-box rules preclude publicly traded entities, “per se” corporations, financial institutions, as

well as certain entities subject to special tax treatment from making the election. The exclusion of

per se corporations was required because of the limited mandate of the regulations, and has no

sound basis in policy. On first principles, all business entities should be allowed to make the

election, with specific exceptions based on valid policy or administrative reasons. Thus, the first

restriction is that the election should be restricted to business entities – for example, a personal

trust80 should not be allowed to elect to be treated as a partnership. There are presumably legitimate

policy reasons why trusts are not allowed to flow out losses and deductions to their beneficiaries,

and an elective classification system should not be permitted to upset this scheme.

Next, publicly traded and other widely held entities would not be allowed to elect their tax

status. Allowing public corporations to elect partnership status would be tantamount to an elective

abolition of the corporate tax. Obviously, this would be unacceptable from a revenue standpoint. It 79 Subdivision j – sections 96-103 of the Act. 80 Defined in subsection 248(1). Essentially, a testamentary trust or a trust no beneficial interest in which was acquired for consideration. The latter definition would cover the typical family trust.

23

would also make the Canadian tax system out of sync with the rest of the world, as the author is not

aware of any country with an income tax that does not tax corporations at the entity level. Finally,

the administrative complexity of dealing with a “partnership” having many thousands of members

and actively traded interests would be significant. It is one thing to reassess a public corporation or

a closely held partnership. It would be a quite different matter to reassess thousands of shareholders

who had no responsibility for the error or omission (or abusive avoidance transaction) and try to

collect taxes, interest, and penalties from them.81 Thus, the elective system should be restricted to

non-public entities (private businesses and subsidiaries or “joint ventures” of public entities).

The election should apply to both domestic and foreign entities, as allowing foreign

affiliates to specify their tax status will reduce uncertainty and complexity in the foreign affiliate

system. Other than the restriction on public trading, there should be no restrictions on the size or

number of members of the eligible entities. Further, the author contends there should be no

restrictions on the types of members of an electing entity. This raises the issue of potential revenue

losses if interests in a conduit entity are held by non-residents or tax exempts. While this is a

legitimate concern, the author does not believe it to be a sufficient justification for denying pass-

through treatment to the other members of the entity. Instead, it would make more sense to simply

impose the regular corporate tax on any income allocated to non-residents or tax-exempts.82 To deal

with the risk that a non-resident member could claim treaty protection based on the argument that it

has no permanent establishment in Canada, a provision should be added to the Income Tax

Conventions Interpretation Act83 to deem any non-resident holding an interest in an entity electing

conduit treatment to have a permanent establishment in Canada.

Finally, there is no reason to limit the election to formal entities. If a proprietorship wants to

“incorporate” for tax purposes or a corporation wants to “incorporate” one of its divisions without

actually creating a corporation, there appears to be little reason to prevent them from doing so. As

explained below, the proposal would make any conversion a liquidation subject to the ordinary tax

consequences of this type of event, which should cut down on opportunistic elections.

81 One could, of course, specify that the “partnership”, rather than the members, would be liable for any additional taxes, interest, and penalties. Nevertheless, this would raise other issues, such as what tax rate to apply to the deficiency (the corporate rate, the top marginal individual rate, or some average rate?). 82 The exact scope of the tax exempts subject to this rule (e.g. whether it should apply only to institutional investors, such as pension funds, or also to RRSPs) would be a policy decision and depend in part on the amount of revenue at stake. 83 R.S.C. 1985, c. I-4.

24

Treatment of Single Member Entities

In the run up to the introduction of the check-the-box rules, there was much discussion in the

US concerning the treatment of single member entities. The issue was whether these entities should

be allowed to elect, since in the original proposal the only alternatives were either corporate or

partnership status. It was felt that allowing single member entities to elect partnership status would

be inappropriate, as a partnership with a single member is a contradiction in terms.84 The solution

finally settled on was to treat such entities as disregarded for tax purposes, in the same manner as an

unincorporated proprietorship or division. This treatment reflects the economic reality of a wholly-

owned subsidiary as a mere extension of the parent’s business and is thus appropriate. It also

reflects the goal of simplicity, since it will not be necessary to attempt to apply the partnership rules

in a context where it would not make sense to do so.

Income and Loss Allocation

The allocation of the income of a partnership to its members currently follows the provisions

of the partnership agreement, or, if no such agreement exists, the provisions of the relevant

Partnership Act.85 A corporation, on the other hand, does not have an inherent mechanism for

allocating income to its shareholders on a current basis, so the legislation will have to provide for

statutory allocation rules. A simple rule based on share ownership is sufficient if there is only a

single class of shares. However, income allocation becomes more complex if there are several

classes of shares with different entitlements to distributions. In order to reduce opportunities for tax

avoidance, income allocation should be closely tied to the economic interest of the member in the

income and assets of the corporation. The US S Corporation rules eschewed this issue by

prohibiting electing corporations from issuing more than one class of stock. However, such a

restrictive rule would put electing corporations at a significant disadvantage relative to “real”

partnerships and hence would not achieve the neutrality objective.

One way of dealing with several classes of shares would be to treat non-participating

dividends (i.e. dividends on preferred shares) as interest, deductible to the electing entity and fully

taxable to the recipient.86 The residual income would then be allocated to the participating shares

84 New York State Bar Association – Tax Section, “Report on the “Check-the-Box” Entity Classification System Proposed in Notice 95-14,” August 30, 1995. 95 Tax Notes Today 173-64, at paragraph IV.A.1. 85 Paragraph 96(1)(f). 86 This option was part of a proposed amendment to the S Corporation rules, which was however never enacted. See Walter D. Schwidetzky, “Is it Time to Give the S Corporation a Proper Burial?,” Virginia Tax Review vol. 15 (1996), at 630.

25

based on their right to participate in profits if all income were paid out currently as a dividend.

Losses would be allocated to the participating shares by deeming the loss to be distributable surplus

and determining the amount of the dividend that would have been received if this notional surplus

had been paid out as a dividend. An allocation system based on distribution entitlement would

obviously be complex, and proper safeguards will be required to prevent abuses in non-arm’s length

situations, such as the use of shares with discretionary dividends to achieve income splitting.

A related issue is whether the current flexibility to allocate partnership income should be

restricted in some way. Paragraph 96(1)(c) provides that the income allocated to a partner is

computed as if each partnership activity (including the ownership of property) were carried on by

the partnership as a separate person. This essentially allows a partnership to allocate income from

different sources in different proportions (i.e. there is no consolidation at the partnership level).

While such special allocations may be undertaken for bona fide business purposes, they also offer a

lot of opportunities for tax planning. While the Act contains rules that allow the Minister to re-

allocate income among members if the principal purposes of an allocation is tax avoidance87 or if

the allocation is between non-arm’s length parties and is not reasonable,88 it is uncertain whether

these general anti-avoidance provisions are a sufficient safeguard, especially if the use of entities

electing partnership status were to become commonplace.

Default Rules and Mechanics of Election

In the absence of any election, the default position for domestic entities would be the current

treatment (i.e. entity level tax for corporations and conduit treatment for partnerships). For foreign

entities, the default position would be (i) partnership treatment if the entity has several members and

at least one member has unlimited liability, (ii) corporate treatment if all members have limited

liability, and (iii) disregarded if the entity has a single owner with unlimited liability. The election

to change the tax status should be made at the entity level, not at the member level. Allowing each

member to elect either corporate or conduit treatment separately from the other members would be

excessively complex and give rise to additional tax planning opportunities. The election should be

by unanimous consent of all the members holding an equity interest in the entity to prevent actions

that are detrimental to minority shareholders and holders of preferred shares.89 An election should

87 Subsection 103(1). 88 Subsection 103(1.1). 89 In the US, an S Corporation election requires unanimous consent by all shareholders. A check-the-box election must be signed either by all the owners of the entity or an officer of the entity who is authorised (either under local law or the

26

only be effective prospectively to prevent retroactive tax planning. Finally, once an affirmative

election to adopt a status different from the default rules is made, no change in status would be

allowed for five years unless there has been a change of control and the Minister is satisfied that a

status change would not result in undue tax avoidance. This restriction on the flexibility of the

system is appropriate to reduce the opportunity for opportunistic elections to take advantage of

specific situations or events making a particular tax treatment more attractive.

Treatment of Conversions

A conversion from one status to another should be treated for tax purposes as if the event

had actually happened. Thus, as under the check-the-box rules, a conversion from corporate status

to partnership or disregarded status would be deemed to be a winding up and liquidation of the

corporation followed (in the case of an entity with more than one member) by a contribution of the

assets and the liabilities to a newly formed partnership in return for an interest in the partnership.

The provisions of subsections 88(1), 88(2), 69(5) dealing with corporate liquidations and subsection

97(1) dealing with contributions to a partnership would apply as in a “real” conversion. The

members may elect a tax-deferred contribution under subsection 97(2) if the conditions of that

provision are met. A conversion from partnership to corporate status would be considered a

contribution of the partnership’s assets to a corporation in return for shares followed immediately

by the liquidation of the partnership and the distribution of the (notional) shares to its members. The

rollover under subsection 85(2) would be available if the members so elect.

Treating the conversion as an ordinary liquidation has several advantages. First, there is no

need to keep track of pre-conversion surplus balances, loss carryforwards, and accrued gains or

losses (the rationale for these rules was explained in the section about the S Corporation). Second, it

is appropriate that taxpayers should bear the full tax consequences of their election. To put it

differently, the elective conversion of a corporation to a partnership or division (or vice versa)

should not be more tax advantageous than an actual conversion would have been. The purpose of

elective classification is to allow taxpayers to structure their businesses without being influenced by

tax considerations, not to provide a fundamentally novel tax treatment.

entity’s organisational documents) to make the election. Thus, the standard for this type of election appears to be unanimity. It is also important to ensure that owners of securities that are otherwise non-voting (i.e. preferred shares) have the right to block an election since a status change would radically alter the tax treatment of their securities (i.e. preferred share dividends would be treated as interest if the entity elects conduit treatment).

27

Anti-Avoidance Rules

Loss Trading

The trading of accrued loss carryforwards would not be facilitated by the elective

classification since a conversion would result in an ordinary liquidation of the loss company. Thus,

the rules will not provide for additional planning opportunities beyond what is already possible

under subsection 88(1) and the current change of control rules should be sufficient to deal with any

abuses. However, the change of control rules should be expanded to apply to partnerships to force

the realisation of accrued losses on a change of control (a partnership does not have loss

carryforwards, but it can effectively transfer losses inherent in assets that have fallen in value).

These losses would then be subject to the change of control restrictions. There is currently a rule

that prevents the importation of losses of certain foreign partnerships when a Canadian resident

becomes a member of such a partnership.90 However, there is no similar rule for other partnerships.

Income Splitting

Elective classification could be used to facilitate certain income splitting arrangements.

Currently, the “kiddie tax” (a special tax on minors at the top marginal personal tax rate) applies to

partnership income only if the income is derived from the provision of property or services to a

related person and certain corporations in which a related individual is a shareholder.91 In contrast,

the kiddie tax applies to all dividends form private corporations. To preserve consistency, all of the

income of a corporation that has elected partnership status should be subject to the kiddie tax if it is

allocated to a minor.

Implications of Elective Entity Classification on the Foreign Affiliate Rules

Simplification of Foreign Entity Classification

Proper entity classification is key to the application of the Canadian foreign affiliate system

as only a foreign entity classified as a corporation may qualify as a foreign affiliate, the definition

around which the Canadian international tax system is built. As mentioned earlier, currently there

are no statutory classification rules for foreign entities and taxpayers must rely on CRA guidance

and very limited jurisprudence. This can cause substantial uncertainty when taxpayers are

confronted with foreign entities for which CRA has not published any guidance, particularly since

90 Subsection 96(8). 91 Definition of “split income” in subsection 120.4(1).

28

the laws of many foreign jurisdictions provide for entities whose legal attributes do not fit neatly

into the Canadian categories of corporation, partnership, or trust. Even where the classification of

the entity is not in doubt, the taxpayer may still have problems if the entity is not classified as a

corporation. As a result of amendments introduced in 1999,92 the foreign affiliate rules now deal

reasonably well with partnerships, although there are still uncertainties.93 However, the rules do not

work at all for foreign trusts, even though the use of business trusts is common in certain

jurisdictions (e.g. the United States for state tax planning and non-tax purposes). An elective entity

classification system would alleviate these problems by providing certainty as to the classification

of the foreign entity for Canadian tax purposes and allowing taxpayers to use non-corporate entities

for local business or tax reasons while still treating them as foreign affiliates for Canadian tax

purposes, hence avoiding the problems caused by the failure of the foreign affiliate system to deal

appropriately with non-corporate entities.

Increased Tax Planning Opportunities

Elective entity classification would increase planning opportunities in the area of

international tax and could lead to potential abuses. In the aftermath of the introduction of the

check-the-box rules in the US, the IRS and Treasury have attacked a number of alleged abuses

exploiting the difference in tax treatment for US and foreign tax purposes, some of which could also

apply to a Canadian elective classification system. Some of these planning opportunities and

potential abuses are briefly reviewed in this section.

Avoidance of Foreign Accrual Property Income

The foreign accrual property income (“FAPI”) rules are designed to protect the Canadian tax

base by taxing on a current basis certain types of income earned by a controlled foreign affiliate of a

Canadian taxpayer.94 The US has a similar system, called the “subpart F” rules. One of the planning

techniques attacked by the US authorities involves the use of hybrid disregarded entities to

circumvent the “same country” exception which permits the exclusion of certain inter-affiliate

92 In particular, section 93.1, which deals with dividends received from a foreign affiliate by a partnership. 93 To mention just one, the definition of “equity percentage” in subsection 95(4) only deals with indirect ownership through corporations. The concept of equity percentage is crucial to the definition of foreign affiliate. Thus, it is not clear whether a foreign corporation held indirectly via a partnership or trust qualifies as a foreign affiliate of the Canadian taxpayer. 94 The most important types of income subject to these rules are income from property (passive income), income from certain transactions that erode the Canadian tax base, as well as certain capital gains. The general inclusion of passive income in FAPI is subject to a number of important exceptions for inter-affiliate transactions. In general, inter-affiliate payments will not be FAPI if the payor can deduct the payments from its active business income (paragraph 95(2)(a)).

29

payments from subpart F income.95 In Canada, by contrast, most exceptions from FAPI do not have

a “same country” requirement. Thus, in general, the use of hybrid disregarded entities should cause

fewer concerns in Canada than in the US.

However, there are two provisions in the FAPI rules that do rely on a same country

exception, so that the use of disregarded entities could lead to abuses. The first provision is

paragraph 95(2)(a.1), which is designed to prevent the use of foreign purchasing corporations in

low-tax jurisdictions to reduce Canadian tax. The provision contains an exception for goods that

were produced in the country in which the affiliate was formed and in which its business is

principally carried on. A potential abusive planning technique would involve a purchasing affiliate

in a high-tax jurisdiction (Country A) setting up a wholly-owned subsidiary in a tax haven that is

treated as an entity for Country A tax purposes, but elects to be disregarded for Canadian tax

purposes (a “hybrid branch”). The hybrid branch would then purchase goods manufactured in

Country A and re-sell them to its parent at a higher price, which in turn would sell them to the

Canadian taxpayer. Since the subsidiary is disregarded, the same country exception would be met

provided the parent affiliate’s business is still considered to be principally carried on in Country A.

A portion of the overall spread would be earned in the tax haven,96 hence lowering the group’s

overall tax burden. Arguably, the policy intent of the same country exception is abused because it

was specifically designed to prevent the use of low-taxed 3rd country purchasing corporations.

Similarly, paragraph 95(2)(l) generally includes in FAPI income from a business principally

involved in trading or dealing in indebtedness, including earning interest income. The rule will not

apply if the debtors are arm’s length persons resident in the country in which the foreign affiliate is

resident and principally carries on its business. A disregarded tax-haven subsidiary could be used in

a manner similar to the situation described above to reduce the foreign tax on the income while still

qualifying for the same country exception.

It is not clear how serious these potential abuses would be in practice. In both cases, there is

a requirement that the business be principally carried on in the particular country for the same

country test to apply. Thus, the CRA may be able to challenge the use of hybrid branches on the

basis that the business is principally carried on by these branches, which would deny the same

country exception to the parent. If this factual test is too difficult to apply, specific anti-avoidance

rules should be considered to prevent these abuses.

95 Daniel S. Miller, “The Strange Materialization of the Tax Nothing,” 20 Tax Notes International 2353 (2000), at 2368. 96 Assuming, of course, that Country A itself does not have a rule similar to paragraph 95(2)(a.1)!

30

Surplus Planning

Next to avoiding FAPI, the principal goal of foreign affiliate planning is the creation of

“exempt surplus.” The reason is that exempt surplus can be repatriated to Canada free of Canadian

tax. Exempt surplus generally includes income from an active business carried on in a “designated

treaty country” (“DTC”)97 by a foreign affiliate resident in a DTC, as well as a all or a portion of

capital gains.

One potential use of disregarded entities is to “transform” a share sale into an asset sale for

surplus purposes. One half of the capital gain from the sale of shares of a foreign affiliate is

included in exempt surplus and the other half in taxable surplus;98 this is so regardless of the

underlying assets of the foreign affiliate. On the other hand, 100% of gains from the sale of assets

used in an active business carried on in a DTC are included in exempt surplus. Thus, from a surplus

perspective, a sale of the assets of an active business is usually preferable to a share sale. By

electing to treat the affiliate as a disregarded entity, a share sale can effectively be converted into an

asset sale without foreign tax consequences. This could be viewed as an abusive result given the

clear scheme of the Act that only 50% of the gain on shares should qualify as exempt surplus. Thus,

an anti-avoidance rule that would ignore for surplus purposes a status change made as part of the

same series of transactions as a subsequent sale may be appropriate.

Another potential area of abuse is the treaty-residence requirement for earnings to be

included in exempt surplus. For example, a group financing company resident in a DTC could set

up hybrid subsidiaries in tax havens to earn interest income from related entities.99 The subsidiaries

would be treated as entities for foreign tax purposes but disregarded for Canadian tax purposes, so

the financing affiliate would be considered to earn the income directly. In order to be included in

exempt surplus, the income must be from a business carried on in a DTC. A business is considered

to be carried on in a country if the foreign affiliate has a permanent establishment situated in that

country.100 Since the tax haven subsidiaries carry on no activity except passively earning interest

income, the taxpayer would argue that the financing affiliate does not have a permanent

establishment in the tax haven since no business is carried on there. Thus, the sole permanent

establishment would be in the DTC, and the interest income would be included in exempt surplus,

97 A country with which Canada has a tax treaty. 98 Taxable surplus is fully taxable when repatriated to Canada, subject to a grossed-up deduction for related foreign taxes. 99 It is assumed that the income would qualify for the FAPI exceptions in paragraph 95(2)(a). 100 Regulation 5906.

31

even though it was not subject to tax in the DTC. To prevent this result, a foreign affiliate should be

deemed to have a permanent establishment in a country if it owns a disregarded entity resident in

that country. This would ensure that the income earned by the tax haven entities would be taxable

surplus since it would be attributable to a permanent establishment in a non-treaty country.

Federal-Provincial Tax Issues

Provincial revenue concerns have been a major stumbling block in previous proposals to

achieve some form of loss consolidation or transfer system among corporate groups and may

similarly cause the provinces to oppose the introduction of an elective entity classification system.

This section will review the major provincial concerns that have been voiced against a loss transfer

system, analyse the similarities and differences of the proposed elective entity classification system

as they affect provincial tax planning, and discuss the amendments required to integrate the elective

classification system into the current rules governing provincial income allocation. This section will

also briefly introduce and evaluate alternative options to address provincial revenue concerns.

Specific Provincial Concerns

As mentioned previously, an elective tax classification system is expected to result in overall

revenue losses primarily because it allows the more effective utilisation of losses that are currently

“trapped” in individual corporations. This will be true both at the federal and, assuming the

provinces follow the federal classification scheme, at the provincial level. However, provincial

revenues could suffer proportionately more than federal revenues if the elective classification

system permits corporations to implement more effective provincial tax planning techniques. With

respect to the 1985 loss transfer proposal, the provinces had three major concerns.101 First, while

loss transfers would generally affect only the timing of tax revenues rather than their absolute

amount at the federal level,102 they could lead to permanent provincial revenue losses if a tax loss

generated by an entity in one province is deducted against income generated by an entity that is

taxable in a different province. Second, the system might create increased planning opportunities to

allocate income and losses to exploit variations in provincial tax rates by ensuring that losses are

deducted against high rate taxable income, while subsequent income will be taxed at lower rates in

101 Nicholas LePan, “Federal and Provincial Issues in the Corporate Loss Transfer Proposal,” in Report of Proceedings of the Thirty-Seventh Tax Conference, 1985 Conference Report (Toronto: Canadian Tax Foundation, 1986), at 13:3 and 13:4. 102 Assuming the losses would not otherwise have expired unused.

32

other provinces. Finally, the provinces were concerned that one province could bear the cost of tax

incentives of another province crystallised in the transferred losses.103

Comparison of Loss Transfer and Elective Classification System

In evaluating whether and to what extent these concerns would apply to an elective entity

classification system, it is important to examine the mechanisms by which the two systems

accomplish the consolidation of income and losses. Under the 1985 loss transfer proposal, each

corporation would continue to compute its income as a separate entity and the only effect of a loss

transfer would be to reduce the non-capital loss of the transferor (and thus presumably increase its

taxable income at some point in the future) and to reduce the taxable income of the transferee. Such

a system would obviously permit the shifting of losses from low-tax to high-tax provinces without

the high-tax province benefiting from the increased future taxable income of the transferor. It is

important to note that this effect can already be achieved under current rules by pushing interest

expense or other expenses into the high-tax entity with the corresponding income being earned in a

lower-tax province.104 However, a loss transfer system would simplify this type of planning by

removing the need to set up financing structures and similar planning techniques which may be

costly to implement and subject to challenge by the tax authorities.

In contrast, an elective classification system would disregard the separate existence of

single-member entities and treat multi-member entities as partnerships. The allocation rules in Part

IV of the Income Tax Regulations and the corresponding provincial rules would then apply to

allocate the income and losses of the electing entities to the different provinces. The income or loss

of the corporation is allocated to each province in which the corporation has a permanent

establishment based on the average of the ratio of the gross revenue associated with that permanent

establishment to the total gross revenue (excluding certain types of passive income) and the ratio of

the salaries and wages of that permanent establishment relative to the total wage expense.105 If the

corporation is a member of a partnership, the gross revenue and salaries and wages of the

partnership are allocated to the corporate partners based on their share of the income or loss of the

103 This apparently assumes that the transferable losses for provincial tax purposes would be computed based on the tax rules of the transferor province. However, it would be more reasonable to assume that the losses would be computed based on federal rules, in which case they should not be affected by specific provincial tax provisions. 104 Similar planning techniques are very common in the international tax arena. They should be even easier to implement in an inter-provincial context given the general absence of provincial transfer pricing or thin capitalisation rules. 105 Regulations 402(3) and 402(5).

33

partnership.106 The CRA has clarified that the income allocation should not be done at the level of

the partnership (with the resulting allocation attributed to all partners), but at the level of each

partner corporation.107 While the federal allocation rules are technically only applicable for the

purpose of determining the portion of a corporation’s income eligible for the 10% federal tax

abatement, they will automatically determine the allocation for provincial tax purposes in the seven

agreeing provinces. Further, the rules in Alberta, Ontario, and Quebec generally have the same

result as the federal rules.108

Compared to a loss transfer system, the elective entity classification system should offer

much less flexibility to specifically shift income or losses to minimise the average provincial tax

rate. Both the income and losses of the electing entity will be allocated to all provinces in which the

resulting “consolidated group” has a permanent establishment based on the overall sales and wages

of the group. While a province will still bear the cost of tax losses generated by a transparent entity

in another province, the resulting revenue cost is allocated to all provinces in which the corporate

group has operations, and the group does not have the same flexibility to shift the loss to the highest

tax province as it would under a loss transfer system. In addition, the province would also share in

the subsequent income, which has been increased by the utilisation of the loss, on the same basis as

it shared in the loss, assuming the distribution of salaries and wages did not change substantially.109

In effect, the provinces should be in the same situation as the federal government: the accelerated

loss utilisation will lead to a timing difference in the collection of revenues, but is less likely to

result in a permanent revenue loss. Thus, an elective classification system should not necessarily

create greater revenue concerns at the provincial level than at the federal level.

A major factor determining the extent of provincial tax planning opportunities is the ease

with which taxpayers can switch between pass-through and entity treatment depending on which

option is more favourable. For example, a taxpayer with an entity in a low-tax province would like

to elect pass-through treatment when the entity has losses (so they can be offset against income

from other provinces taxed at higher rates) while electing separate entity treatment when the entity

has positive income (so that all income is taxed at the low rate instead of being allocated in part to

higher rate provinces). If the taxpayer were permitted to make the election annually on an ex post

basis, this type of planning would be easy and the elective entity classification regime would be

106 Regulation 402(6). 107 CRA, Technical Interpretation no. 2000-0054815 (E) Allocation of Partnership Income to Provinces. 108 See, for example, Ontario Corporate Tax Regulations 302(3) and (6). 109 Supra, note 101, at 13:9.

34

similar to a loss transfer system. If, on the other hand, the election must be made ex ante and can

only be changed on an infrequent basis, if at all, this type of one-sided loss shifting can be virtually

eliminated. Thus, as mentioned earlier, taxpayers should be required to maintain their elected status

for a minimum number of taxation years in order to reduce the potential for opportunistic changes.

A period of five years was suggested above because this would be consistent with similar systems

both in Canada (pre-1952 group consolidation) and the US (check-the-box rules). However, five

years may be too short to prevent provincial tax planning, especially in start-up situations, so a 10

year term may be preferable. In addition, anti-avoidance rules will be required to prevent taxpayers

from circumventing the limitation period. Thus, for example, if an electing entity transfers all or

substantially all of its assets to a new entity controlled by the same taxpayer(s), the new entity

should be deemed to be a continuation of the transferor for the purposes of the limitation period.110

If these measures are taken, provincial tax planning opportunities under the elective classification

system should not be substantially increased relative to their current level.

Required Changes to the Income Allocation Rules

While the current income allocation rules should in general be appropriate to deal with the

proposed elective entity classification system, certain amendments will be required to allow them to

deal with entities that have a separate legal existence (i.e. corporations). While the current

regulations incorporate a rule to allocate the revenue and salaries & wages of a partnership to its

corporate partners, there is no explicit rule that attributes a permanent establishment of the

partnership to its partners. Nevertheless, it is the CRA’s position that each member of a partnership

has a permanent establishment in each province where the partnership has a permanent

establishment. This applies to both general and limited partners.111 Apparently, this position is

based on the case of No. 630 v. MNR,112 in which the court held that, since a partnership is not a

separate (legal) entity, it would be logically impossible for the partnership to have a permanent

establishment that is not, at the same time, the permanent establishment of its partners. Since a

corporation is a separate legal entity, this reasoning would not apply to a corporation that is treated 110 The author would allow an exception if the transfer was part of a series of transactions resulting in a change of the control of the entity. This is to allow taxpayers to reincorporate disregarded entities prior to a sale to achieve the tax results of a share sale instead of an asset sale. 111 See, for example, CRA Technical Interpretation no. 2000-0048685 (E) Allocation Taxable Income Between Provinces. 112 59 DTC 300, at 303 (Tax Appeal Board), cited in Richard G. Tremblay, "Permanent Establishments in Canada," in Report of Proceedings of Forty-First Tax Conference, 1989 Conference Report (Toronto: Canadian Tax Foundation, 1990), 38:1-69.

35

as a pass-through entity under an elective classification system. Accordingly, a specific deeming

rule must be introduced into the federal and provincial income allocation regulations to provide that

any permanent establishment of an entity that elects to be taxed as a conduit is deemed to be a

permanent establishment of each of its shareholders or members. With this amendment, the income

allocation rules should work appropriately under an elective entity classification system.

Alternative Options to Address Provincial Revenue Concerns

As explained in the previous section, an elective entity classification system should cause

fewer provincial revenue concerns than the failed 1985 loss transfer proposal. Nevertheless, it is

likely that the provinces would be uncomfortable with adopting elective entity classification, be it

only because of the uncertainty of its consequences. Thus, this section will briefly explore other

alternatives to protect the provincial revenue base.

Provincial Non-Conformity

Some provinces may decide not to follow the federal classification scheme and instead

retain entity taxation for all corporations. This option is undesirable since it would remove a portion

of the benefits of elective classification and increase the complexity of the tax system. If the

provinces do not follow the federal election, the objective of achieving neutrality among different

business forms will not be achieved. If flow-through treatment cannot be achieved for provincial

purposes via the election mechanism, the current incentives to achieve de facto consolidation

through business combinations and financing arrangements will remain, and the elective

classification system will add another layer of complexity onto an already complex tax structure.

More seriously, a federal-only system could lead to a complete divorce between the federal and

provincial tax systems. Not only would federal taxable income differ substantially from provincial

taxable income, but entities that do not even have a federal filing obligation would be required to

file provincial corporate tax returns. For example, a corporation with a large number of wholly-

owned subsidiaries might file just a single federal return (assuming it elects disregarded status for

all its subsidiaries), but each one of these subsidiaries, as well as the parent, would have to file a

separate provincial corporate tax return.

Under these conditions, it is likely that non-conforming provinces could no longer be part of

the federal-provincial Tax Collection Agreements because the differences between federal and

provincial taxable income would be too large to adjust for. This factor alone might put significant

pressure on the agreeing provinces to conform to the federal system, since for many of these

36

provinces the creation of a separate corporate tax administration would be too costly. Conversely,

this pressure would likely lead to increased provincial resistance to a federal elective classification

system since the provinces would not have the alternative to simply ignore it. Therefore, a federal-

only system is not only undesirable, but likely also impractical, and better alternatives should be

sought to address provincial revenue concerns.

Inter-Provincial Compensation Arrangements

In the debate around the 1985 loss transfer proposal, it was suggested that compensating

payments could be used to offset provincial revenue losses. The province of the loss transferor

would make a payment to the transferee province in an amount equal to the (expected) tax value of

the loss. The discussion, however, pointed out the difficulties of determining an appropriate

compensation amount, especially where the transferor and/or the transferee corporations are taxable

in more than one province.113 The measurement problem would be even more severe in an elective

entity classification system, in which there are no obvious “loss transfers” but rather a consolidation

of all income and losses of the electing entities and their shareholder(s). In addition, determining the

amount of the compensation requires an assumption about the distribution of tax liabilities in the

absence of the election, which would be arbitrary since the election presumably affects the

taxpayer’s behaviour.114 Lastly, a compensation system might be viewed as an indirect tax by a

province on income earned in another province and may be ultra vires under the Constitution.115

For all of these reasons, a loss compensation system, either bilaterally between the provinces or

administered by the federal government, is not a feasible solution.

Surtaxes and Alternative Taxes on Electing Entities

One possible alternative would be for the federal government to follow the example of the

pre-1952 consolidation regime and levy a surtax on the income allocated from entities electing

conduit treatment. The proceeds of the surtax would then be distributed to the provinces as

compensation for their revenue losses. However, this alternative suffers from several weaknesses.

First, it is not obvious how the income that should be subject to the surtax is to be measured. One

alternative would be to include the corporation’s share of the income of electing entities that would

have been taxed as corporations in the absence of the election. However, this would require

corporations to keep track separately of the income of otherwise “disregarded” entities (i.e., wholly-

113 Supra, note 101, at 13:9. 114 I.e., in the absence of the election, the taxpayer would likely have structured its expenses differently to utilise the losses. 115Supra, note 101, at 13:11.

37

owned subsidiaries) and would thus increase the complexity of the system. It is also not clear that

the income of the electing entities is the appropriate measure of the tax benefit obtained by the

corporation. Second, this proposal would generate the same difficulties as the compensating

payments in determining an appropriate distribution of the surtax proceeds to the provinces. An

allocation based on the amount of taxable income allocated to a particular province may bear little

relationship to the actual losses suffered by that province, and would even “reward” provinces that

obtain a net revenue gain from the elective tax classification system.116 Thus, a surtax would not be

a conceptually desirable solution to provincial revenue concerns, although it could be a useful

political expedient to make elective entity classification acceptable.

Another alternative would be for the provinces to levy business taxes that do not depend on

the status of the entity as a corporation. Such taxes would most likely be based on some measure

other than income. Possible alternatives would include a gross receipts tax on all businesses and

increased filing fees for entities that elect pass-through treatment. Each of these alternatives has its

own drawbacks. A gross income tax is similar to a sales tax and would have to be coordinated with

provincial sales taxes in order to prevent creating an excessive tax burden for certain businesses. In

addition, a gross receipts tax is inequitable because it does not take into account the profitability of

the business.117 An example of a filing fee that is in essence a tax on a particular entity type is the

$4,000 tax on incorporation118 and $2,000 annual filing fee119 for a Nova Scotia Unlimited Liability

Company (“NSULC”)120 enacted by the Nova Scotia government in 2002.121 It would be similarly

possible to impose higher filing fees for corporations that elect to be taxed as pass-through entities

in an elective classification system. The advantage of filing fees is their simplicity and ease of

administration. However, filing fees do not take into account the size or profitability of the business

and thus are highly regressive in nature. From a tax policy perspective, they are not an appropriate

116 Since elective tax classification does not change the overall taxable income across all corporations, if some provinces lose revenue due to the proposal, other provinces necessarily must gain revenue. However, it would be difficult to determine which provinces are net winners or losers because there is no clear basis for comparison. 117 Pommy Ketema, “Did the Federal Check-the-Box Regulations Open Up a State Tax Pandora’s Box? A Reflection on State Conformity to the New Federal Classification Scheme of Single-Member LLCs,” 82 Minnesota Law Review 1659 (1998), at 1690. 118 Subsection 5(2) of the Nova Scotia Companies Act, R.S.N.S., c. 81, s. 1. 119 Nova Scotia Registry of Joint Stock Companies. Notice on Registry of Joint Stock Companies Fee Increases, dated April 1st, 2004. Available on the Web at <http://www.gov.ns.ca/snsmr/pdf/rjsc/RJSC_Notice_and_Fee_Schedules_april1_04.pdf>. 120 Most NSULCs are used by US corporations to create hybrid structures since the NSULC is the only corporate form in Canada that is an eligible entity under the check-the-box rules. Typically, these entities file their tax returns on the basis that they have no permanent establishment in Nova Scotia and thus pay neither Nova Scotia income nor capital taxes. The filing fee is often the only way Nova Scotia will generate any revenue from these entities. 121 For comparison, the corresponding fees for a “regular” limited company are $270 and $95, respectively.

38

replacement for an income tax. In addition, it would be difficult to establish an optimal flat filing

fee that generates sufficient revenues without unduly discouraging small businesses from taking

advantage of the elective classification rules. To partly address this issue, it would be possible to

index the fees wholly or partly to the size of the business, but that would increase the complexity of

their administration.122 Thus, while filing fees may serve as an additional source of provincial

revenue, they are not a substitute for the taxation of corporate income.

Conclusion

Over the past few years, there have been renewed calls for Canada to depart from its strict

separate entity taxation of corporations. The 1997 Mintz Committee Report recommended that the

federal government consider developing, in consultation with the provinces, a formal system for

transferring losses between members of the same corporate group.123 More recently, the Canadian

Chamber of Commerce124 and the Tax Executives Institute125 have also called for the speedy

adoption of a formal loss transfer system, stating that the absence of group relief mechanisms may

stifle innovation and hurt Canada’s global competitiveness. Concurrently, we have witnessed the

proliferation of income trusts and other structures taking advantage of the differential tax treatment

accorded to non-corporate vehicles, even if their economic substance is equivalent to that of a

traditional public corporation. Thus, the current Canadian approach of basing the taxation of

business entities on their legal form is no longer appropriate and serves neither the interests of

business nor those of government.

In this paper, the author has proposed to use an elective entity classification system to unlink

the tax treatment of businesses from they way they are structured. An elective classification system

will provide for an efficient means of consolidating losses within a corporate group without having

to introduce a complex set of consolidation rules. Compared to a loss transfer system, an elective

classification system should create fewer provincial revenue concerns because it offers less

flexibility to shift losses between provinces and ensures a certain degree of parallelism between the

loss transfer and subsequent income. An elective classification system will also permit the owners

122 Supra, note 117, at 1690. 123 Canada, Report of the Technical Committee on Business Taxation (Ottawa: Department of Finance, April 1998), at page 4.18. 124 The Canadian Chamber of Commerce. Positions on Selected National and International Issues 2002, 2003 and 2004, at page 72. Available on the Web at <http://www.chamber.ca/cmslib/general/policybook2004.pdf>. 125 Tax Executives Institute. Implementation of a Formal Loss Transfer System. June 20, 2003. Available on the Web at <http://test.tei.org/Resource.phx/public/NEWTEI/include.htx?file=losstransfer.html>.

39

of start-up businesses to personally use the losses of the business while at the same time enjoying

the liability protection of the corporate form. Elective classification would therefore increase the

incentives for risk-taking and innovation. Finally, an elective classification system will improve the

simplicity and certainty of the application of Canada’s foreign affiliate rules. Thus, elective entity

classification is the basis for achieving a rational system of taxing business entities and should be

adopted in Canada.

40

Bibliography Berger, Curtis J., “W(h)ither Partnership Taxation?,” Tax Law Review (New York University

School of Law), vol. 47 (1991), at 105. Canada, Department of Finance, Budget Papers, A Corporate Loss Transfer System for Canada,

May 23, 1985. Canada, Report of the Technical Committee on Business Taxation (Ottawa: Department of Finance,

April 1998). Canada Revenue Agency, Interpretation Bulletin IT-343R - Meaning of the term corporation, dated

September 26, 1977. Canada Revenue Agency, Technical Interpretation no. 2000-0048685 (E) - Allocation Taxable

Income Between Provinces, dated October 20, 2000. Canada Revenue Agency. Technical Interpretation no. 2000-0054815 (E) - Allocation of

Partnership Income to Provinces, dated November 17, 2000. Couzin, Robert, “Income Taxation of Groups of Corporations: The Case for Consolidation,” in

Report of Proceedings of the Thirty-Sixth Tax Conference, 1984 Conference Report (Toronto: Canadian Tax Foundation, 1985).

Cronkwright, Glen E., “The Utilization of Losses in Corporate Groups and Further Relief that

Might be Taken,” in Report of Proceedings of the Thirty-First Tax Conference, 1979 Conference Report (Toronto: Canadian Tax Foundation, 1980).

Fleischer, Victor E., “If it Looks Like a Duck: Corporate Resemblance and Check-the-Box Elective

Tax Classification,” 1996 Columbia Law Review 518. Horwich, Richard J., A Comparative Study of Consolidated Returns and Other Approaches to the

Multiple Corporations Problem, Tax Law Review (New York University School of Law), vol. 20 (1964-65).

Income Tax Act, R.S.C. 1985, c. 1 (5th supplement), as amended to July 2004. Internal Revenue Code of 1986, as amended. Ketema, Pommy, “Did the Federal Check-the-Box Regulations Open Up a State Tax Pandora’s

Box? A Reflection on State Conformity to the New Federal Classification Scheme of Single- Member LLCs,” 82 Minnesota Law Review 1659 (1998).

Lanthier, Allan R. and Plutte, Kerry L., “International Hybrids: Pitfalls and Practice,” in Report of

Proceedings of the Fifty-First Tax Conference, 1999 Conference Report (Toronto: Canadian Tax Foundation, 2000).

41

LePan, Nicholas, “Federal and Provincial Issues in the Corporate Loss Transfer Proposal,” in Report of Proceedings of the Thirty-Seventh Tax Conference, 1985 Conference Report (Toronto: Canadian Tax Foundation, 1986).

Miller, Daniel S., “The Strange Materialization of the Tax Nothing,” 20 Tax Notes International

2353 (2000). New York State Bar Association – Tax Section, “Report on the “Check-the-Box” Entity

Classification System Proposed in Notice 95-14,” August 30, 1995. 95 Tax Notes Today 173-64.

Nova Scotia Companies Act, R.S.N.S., c. 81, s. 1. Nova Scotia Registry of Joint Stock Companies. Notice on Registry of Joint Stock Companies Fee

Increases, dated April 1st, 2004. Available on the Web at <http://www.gov.ns.ca/snsmr/pdf/rjsc/RJSC_Notice_and_Fee_Schedules_april1_04.pdf>.

Omer H. Patrick v. MNR, 1 DTC 303 (Exchequer Court, 1935) Ontario, Corporations Tax Regulation 183, R.R.O. 1990, as amended. Richardson, Stephen R., “A Corporate Loss Transfer System for Canada: Analysis of Proposals,” in

Report of Proceedings of the Thirty-Seventh Tax Conference, 1985 Conference Report (Toronto: Canadian Tax Foundation, 1986).

Richardson, Stephen R., “Transfer of Deductions, Credits, or Losses within Corporate Groups: A

Department of Finance Perspective,” in Report of Proceedings of the Thirty-Sixth Tax Conference, 1984 Conference Report (Toronto: Canadian Tax Foundation, 1985).

Schwidetzky, Walter D., “Is it Time to Give the S Corporation a Proper Burial?,” Virginia Tax

Review vol. 15 (1996), at 591. Stikeman Income Tax Act Annotated 1949. Richard De Boo Ltd. Stikeman Income Tax and Excess Profits Tax. Consolidated Acts 1944-45 Annotated. Richard De

Boo Ltd. Tax Executives Institute. Implementation of a Formal Loss Transfer System. June 20, 2003.

Available on the Web at <http://test.tei.org/Resource.phx/public/NEWTEI/include.htx?file=losstransfer.html>.

The Canadian Chamber of Commerce. Positions on Selected National and International Issues

2002, 2003 and 2004. Available on the Web at <http://www.chamber.ca/cmslib/general/policybook2004.pdf>.

Tremblay, Richard G., "Permanent Establishments in Canada," in Report of Proceedings of Forty-

First Tax Conference, 1989 Conference Report (Toronto: Canadian Tax Foundation, 1990), 38:1-69.

42

Yin, George K., “The Taxation of Private Business Enterprises: Some Policy Questions Stimulated by the “Check-the-Box” Regulations,” Southern Methodist University Law Review vol. 51 (1997), at 125.