10 October 2000
National Treasuryís Response To The Hearings Regarding
The Proposed System of Worldwide Taxation
Dear Madam Chair,
Thank you for affording us with this opportunity to respond to the comments made at the October 6th and October 10th hearings. We at National Treasury and SARS will fully incorporate these comments into the proposed legislation to the extent these comments do not undermine our core principles. In making our responses, we at National Treasury plan to address the general policy issues raised. SARS will respond to the technical detail and administrative issues.
I. FAIRNESS REQUIRES THAT WE WIDEN THE TAX BASE
At the outset, I would like to reiterate the intended purpose of our move from a "source plus" system of taxation to a "residence minus" system.
II. CALLS FOR "INTERNATIONAL COMPETITIVENESS" AND OTHER EFFORTS TO PREVENT THE PROPOSED RESIDENCE SYSTEM
Certain high-profile taxpayers have consistently attacked our proposed move to a residence system. Their main plea is their call to maintain "international competitiveness." They are now mounting an additional attack on the grounds of process, criticising the rapidity of the process and contending the process lacked full consultation.
* * * *
We fully understand that a South African tax on South African controlled foreign businesses potentially impacts international competitiveness. Indeed, our government has gradually relaxed its foreign exchange controls to increase this competitiveness. However, many rightly suggest that an immediate 30-percent South African tax (before foreign tax credits) creates an added burden for South African controlled foreign businesses when those businesses would otherwise be subject to a lower foreign rate. What these critics fail to point out is that the taxation of South African controlled foreign businesses involves balancing of a variety of factors.
* * * *
With respect to criticisms about process, we realise that the proposed legislation has moved rapidly. However, we believe this rapidity is necessary to prevent an unwarranted flight of capital from our tax base. This loss of capital has a strong adverse impact on our efforts to promote internal growth and creates an unfair burden on the domestic salaried classes remaining behind.
Before turning to specific policy issues, we must note that the proposed residence system is highly reasonable given the options available. We chose to target only a limited set of South African controlled foreign business activity that represents a direct threat to the South African tax base. We chose not to go as far as many systems, such as the United States and certain OECD countries, which additionally target all tax haven transactions, even when the tax base of their own countries is not at stake. Although we initially leaned in this direction, we instead chose a more moderate path that fully takes into account the sensitivities of the South African economy. If we ultimately concede to some of the demands now before us, these concessions may unfortunately leave us in the same position as the limited "source plus" system currently in place.
III. ISSUES RELATING TO INDIVIDUALS
Although several issues exist involving individual taxpayers with foreign earnings, I leave most of the detailed aspects, such as the technical wording and administration of the new resident definition, to our counterparts in SARS. I would instead like to address the two most contentious issues raised at the hearings: (1) the issue of foreign wealthy retirees, and (2) the issue of South African residents working abroad.
1. Foreign Wealthy Retirees
With the advent of our residence system, foreign wealthy retirees will become fully subject to tax on their foreign investments just like the remaining South African population. Some have argued that this proposed tax would drive away new retirees who can seek a better deal elsewhere. The proposed tax will also allegedly induce current retirees to leave or simply evade the tax altogether. In turn, this loss of retirees will mean the loss of investment in South Africa, especially in the Western Cape. We at the National Treasury strongly believe these concerns are overstated.
Much ado has additionally been made of the fact that foreign wealthy retirees will form new businesses within South Africa. We again see no statistical evidence. We further question whether this sector of individuals would come to South Africa in their golden years with the inherent desire to start a new business. These individuals will more likely arrive with their stated intent Ė to retire. As such, they will probably not create high quality jobs for the South African economy. Also, if these retirees do in fact start South African businesses, the resulting South African business income will be subject to South African tax regardless of the proposed legislation.
Having said this, we again reiterate the importance of horizontal fairness. Why should we exempt foreign wealthy retirees without similarly exempting our own retirees who have remained within South Africa all of their lives? Both individuals fully utilise South African services and local resources. In that vein, the statistics presented today fail to show the cost borne by the South African government (i.e., in effect borne by other South African taxpayers) in terms of services incurred for hosting these foreign wealthy retirees. These costs are not without significance. All we ask is that foreign wealthy retirees bear their fair share of the cost.
In sum, we are not trying to chase away foreign capital brought by foreign wealthy retirees. We instead believe the best way to attract capital to our shores is by lowering the rate by broadening the base. We do not believe capital is ultimately attracted through selective tax incentives.
2. South Africans Working Abroad
The proposed residence system will create additional tax for South Africans working abroad. Under the new regime, South African residents working abroad will be fully subject to tax on their foreign service and other worldwide income. However, the proposed system contains a generous exemption for foreign service income to the extent South African residents continuously work abroad for 183 days or more (i.e., more than half a year). This exemption exists because these South African residents fail to fully utilise South African governmental services and infrastructure during their absence. At issue is the scope of this exemption.
The speakers today urge us to further expand the exemption. As with the issue of foreign wealthy retirees, we again hear predictions of dire consequences. This proposed tax would allegedly increase the cost of employing South African workers, thereby making South African multinationals non-competitive and South African workers less employable. Although we are amenable to technical changes (such as a de minimis exception to the continuous work requirement), we are reluctant to embrace any effort that greatly expands the exemption.
As a side note, some speakers claim the proposed tax is unfair because many contracts were negotiated before the residence system was proposed (thereby necessitating a delayed effective date). We again reject this argument. These workers have had nearly a year to adjust to the proposed tax. More importantly, all taxpayers subject to any new tax suffer this same burden to the extent they were previously committed to long-term contracts. It should be remembered that tax policy is always subject to reform. We are currently engaged in a reform process that has seen enormous gains that accrue to South Africa as a whole. We should not yield to cries for "fiscal stability clauses" to the detriment of South Africa in favour of a few.
Again, when it comes to the issue of fairness, we reiterate the importance of broadening the base. We want all South Africans to be taxed equally so that all South Africans can share in a lower rate. Selective exemptions mean higher rates for the remainder. Thus, South Africans working abroad should be subject to the same level of tax as those working at home, except to the extent those working abroad have generally foregone the use of South African government services and infrastructure. South African workers who depart for minor periods (usually leaving their family behind) have chosen not to forego these benefits.
IV. ISSUES RELATING TO CORPORATIONS
A. Foreign Income Subject to Tax Under Proposed Section 9D*
The heart of the proposed residence system lies in its provisions dealing with South African multinationals. Although the issues involving individuals have strong emotional appeal, the real tax money at stake is generated by South African controlled businesses operating abroad.
Under the system as proposed, no additional South African tax will apply to foreign branch income or to controlled foreign subsidiary income if: (1) that income is subject to a foreign statutory rate of 27 percent or more (i.e., a rate comparable or higher than the South African rate), and (2) foreign determination of that income is made on a substantially similar basis to that made by the South Africa tax system. The proposed system ignores this income because international law demands that credits be allowed against foreign taxes, and these required credits eliminate almost all South African tax otherwise due when the foreign rate comes close to the South African rate. Largely at issue is how the proposed residence system will tax operations within low-taxed countries (i.e., with rates below the 27 percent systems previously described).
With respect to low-taxed operations, the proposed residence system will fully tax foreign branch income and create partial immediate taxation with respect to controlled foreign subsidiaries (i.e., foreign subsidiaries that are more than 50 percent owned by South Africans). Under current law as introduced by section 9E in July 2000, South African tax does not apply to the income of a controlled foreign subsidiary until that income is repatriated to South Africa (an event that often never occurs). However, under the proposed residence system, a stricter level of tax will apply to:
Under this proposed stricter system of taxation, South African tax will instead apply to controlled foreign subsidiary income as it is immediately earned (rather than at the point of dividend repatriation).
The purpose of this system of partial immediate taxation is to balance the desire for horizontal fairness with international competitiveness. As previously stated, horizontal fairness demands that South African controlled foreign businesses be subject to the same level of tax as South African domestic businesses so that artificial incentives do not exist to direct South African capital away from our shores. International competitiveness, on the other hand, conversely demands that the South African rate not exceed the foreign local rate.
Example. Facts. South African Parent owns all the shares in Foreign Subsidiary located in Country X. Country X imposes a 10 percent tax rate on income arising within Country X. Foreign Subsidiary earns R40 million within Country X.
Result. If the South African tax system ignores the R40 million of income, the R40 million will be subject to a 10 percent rate, whereas South African income is subject to a 30 percent tax rate (thereby violating horizontal fairness). If the South African tax system imposes tax, the R40 million will be subject to a 30 percent tax rate when Country X local rivals will be subject only to a 10 percent rate (thereby violating international competitiveness).
Given this internal conflict, we chose a targeted approach that imposes immediate South African tax only in the following selective situations:
Category #1 - Mobile Foreign Passive Income: South Africa, like most capital exporting countries, currently taxes mobile foreign passive income (e.g., dividends from foreign portfolio stocks and interest from foreign portfolio bonds). Failure to tax these forms of mobile foreign passive income would mean that South African multinationals could easily avoid tax on their passive investments by having their foreign subsidiaries purchase foreign shares and foreign bonds on their behalf. International competitiveness is not at issue in these circumstances because no business activities are involved. Under the proposed residence system, taxation of this category will be extended. For instance, South African controlled foreign businesses can no longer escape the ambit of the tax merely by contending that the passive income is connected to a "substantive business" (a showing that tax experts can satisfy in short order).
Category #2 - Mobile Foreign Business Income: One of the new targets of the proposed residence system are businesses that can easily be located anywhere (including South Africa) but have largely been located abroad for tax savings. In todayís growing world of global electronic commerce, this easy shift of operations to the lowest tax location has become of increased significance. Taxpayers can avoid the ambit of this category of income only by engaging in a foreign business that qualifies as a "business establishment." A business establishment exists if the foreign business involved has some permanence, some hard assets, and a non-tax business reason for operating abroad rather than at home.
Category #3 - Diversionary Foreign Business Income: Another new target of the proposed residence system involves diversionary foreign business income. This category essentially targets situations that will most likely lead to transfer pricing, which artificially shifts South African taxable income offshore. For instance, these circumstances often arise when a South African company purchases products from, or sells products to, a wholly owned foreign subsidiary because the related party price may be over-inflated or under-inflated. Related party services raise similar concerns. The proposed residence system essentially acts as a backstop to transfer pricing Ė a problem that is presently very difficult to enforce.
The speakers as well as the written submissions mainly direct their concerns at our treatment of diversionary transactions. They also request specific exemptions for holding/finance subsidiaries and for relaxation of the reporting requirements. We believe that the proposed legislation can be changed with respect to holding/finance subsidiaries to the extent these changes do not undermine our targeting of the three categories. SARS will deal with the issue of reporting, which implicates their ability to enforce the proposed legislation. However, we do not believe we can accommodate the bulk of the concerns directed toward Category #3, which contains much of the real teeth contained in the proposed legislation.
2. The Importance of Category #3 (Diversionary Foreign Business Income)
Nature of the Problem
As previously stated, the proposed legislation targets diversionary transactions that artificially shift South African taxable income offshore. Diversionary transactions essentially involve sales and service transactions between South African companies and their controlled foreign subsidiaries. The concern in this area is transfer pricing.
Example. Facts. South African Parent owns all the stock of Foreign Subsidiary located in Tax Haven. South African Parent maintains a factory within South Africa that produces gym equipment, costing R200 in parts and labour per machine. South African Parent sells each machine to Foreign Subsidiary for R250. After performing minor packaging and assembly, Foreign Subsidiary resells the gym equipment to independent distributors located throughout the world for R1,000 per machine.
Result. South African Parent reports R50 of taxable income with respect to the gym equipment, even though South African Parent creates almost all the value. Foreign Subsidiary instead reports R750 of the taxable income, which is subject only to Tax Haven tax rates. Both companies are indifferent to this shift as an economic matter because both companies are part of the same economic group.
SARS theoretically has the power under section 31 to reallocate prices so that the above price between South African Parent and Foreign Subsidiary reflects an armís length standard. The problem with section 31 is that the theory fails to match reality. The section 31 power to reallocate prices is extremely difficult to enforce. This problem is not a SARS problem but an international one. Revenue agents from all over the world face the same difficulty. How does one determine an armís length price when this price is absent? While revenue agents have expertise in terms of monitoring the tax impact of transactions, revenue agents do not typically have the significant economics background required to determine price, especially where markets are thin or non-existent. These problems of enforcement are further exacerbated by recalcitrant taxpayers, who often respond by fighting information requests at every turn or by flooding revenue agents with excess information. To make matters still worse, South African multinationals can hire an army of armís length pricing experts who can readily justify all but the most unreasonable of pricing structures.
Impact of the Proposed Legislation
The proposed legislation targets diversionary transactions by acting as a backstop to section 31 when South African companies enter into transactions with related foreign subsidiaries. The proposed legislation accomplishes this backstop result through the following two-fold mechanism:
(1) The proposed legislation provides a new penalty for violating section 31 transfer pricing. Under current law, SARS can only readjust the price to an armís length standard (the price that should have been utilised in the first place). Under the proposal, transfer pricing violations will create South African tax for all the income earned by the controlled foreign subsidiary, even if that income can be supported by section 31 transfer pricing.
Speaker Concerns and Response
The speakers essentially seek to undermine our efforts to prevent diversionary transactions by calling for the removal of the higher business activity standard. They argue that section 31 should be the sole sword against transfer pricing. After all, if the price is legitimate, why should controlled foreign subsidiaries be required to additionally satisfy the higher business activity standard? They further argue this business activity standard is too strict, adversely impacting legitimate business operations. Accordingly, imposition of this standard is again said to upset international competitiveness.
Despite the just described arguments to the contrary, we continue to strongly believe that the higher business activity standard for diversionary transactions should be retained.
We resist current efforts to remove the higher business activity standard in favour of the new transfer pricing penalty. While the new penalty adds some theoretical teeth to section 31, the new penalty does little in terms of practical enforcement. SARS revenue agents must still determine what constitutes an armís length price with all its trappings. In essence, by calling for sole reliance on section 31 with the new penalty, the speakers ask that South African multinationals continue as before.
Some speakers alternatively suggest the proposed legislation target diversionary transactions through sole reliance on the business establishment test. The business establishment test is designed to target shell businesses and other foreign mobile business operations (Category #2). As such, the test is fairly light. All that taxpayers need do is demonstrate that the business involved has minimal substance and a non-tax business reason for operating abroad. The test is insufficient to protect the tax base against transfer pricing when a controlled foreign subsidiary enters into a transaction with a related South African company.
In sum, we are proposing the higher business activity standard as an additional weapon against transfer pricing. While we are ultimately committed toward greater transfer pricing enforcement, we must concede that the technical expertise needed for comprehensive enforcement is still in process. This lack of technical expertise is not unusual. The United States enacted its transfer pricing rules in 1918. Comprehensive enforcement of the U.S. transfer pricing rules only began in the early 1980ís and 1990ís, some 60-to-70 years after enactment.
Many of the speakers and written submissions are requesting a headquarters exception. Under this exception, a South African headquarters company would be exempt from tax on its foreign connected income. As such, a headquarters company could generate foreign income, receive foreign dividends, and distribute dividends to foreign shareholders free of South African tax. In order to qualify for this exception, a South African company would have to be foreign owned and engage almost entirely in foreign activities. This exception presumably will facilitate South Africaís efforts to act as hub for the Southern African region because foreign investors could freely invest in a South African headquarters operation that overseas African operations.
We are willing to consider the proposal but suggest consideration be temporarily delayed until the broader outlines of the residence system are established. Having said this, we do note that the headquarters company exception contains some issues of concern not fully addressed in the Katz Report.
In light of these issues, we are currently reviewing other alternatives in conjunction with the reserve bank. One such option would be an exception for co-ordination centres. It is hoped that a co-ordination centre regime would have more economic substance than a headquarters company while not running afoul of the OECD and European Union efforts to prevent harmful tax competition.
In July 2000, we at Parliament, National Treasury, and SARS spent considerable effort enacting section 9E, which imposes South African tax on foreign dividends for the first time. We have fully debated this issue with exceptions considered. Some of the speakers are now using the proposed residence legislation to revisit section 9E.
1. Foreign Tax Credits and the STC
The speakers argue that we must allow a foreign tax credit against the STC because we have consistently treated the STC as part of our regular corporate income tax for treaty purposes. To allow a credit against the normal tax and not the STC would seemingly violate the principle that the two are one and the same.
The problem with these arguments is that they fail to account for the structure of the proposed legislation as a whole. Because taxation of foreign income always raises multiple complexities, the proposed legislation contains a number of simplifying solutions that cut both for and against taxpayers.
In the case of foreign tax credits, we specifically resist a system of purity. Under a pure residence system, foreign tax credits are subject to multiple basket limitations in order to prevent mixing of high-taxed foreign income against low-taxed foreign income. For instance, the United States subject its credits to a nine basket limitation and other countries utilise a per country limitation. Implementation of these limitations involves complex regulation and compliance. We instead chose a system that generally exempts income subject to a foreign tax rate of 27 percent or more, thereby leaving mostly low-taxed foreign income within the system. As a result of these exemptions, we are proposing a single foreign basket limitation rather than the multiple basket systems used in other countries. In other words, we are allowing South African taxpayers to mix their offshore credits regardless of their foreign location and regardless of the foreign activity involved.
We similarly eliminated use of the foreign tax credit against the STC on grounds of complexity, choosing a 7-year carry-forward instead. Application of the foreign tax credit against the STC creates a whole new set of complications that renders the foreign tax credit provisions untenable. We also note that the case for crediting foreign taxes against the STC is overstated:
2. Foreign Losses
The speakers and written submissions also request liberalisation of foreign losses. Under the legislation as proposed, foreign losses cannot be used to offset South African source income, and losses of one controlled foreign subsidiary cannot be used as an offset against the income of another controlled foreign subsidiary. These speakers and written submissions argue that we must fully allow foreign sources losses to offset South African source income because once foreign income becomes fully part of the tax base so should foreign loss.
We have fully considered these concerns and continue to adhere to our position. Foreign losses should not be available against South African source income and foreign losses of one controlled foreign subsidiary should not be allowed as an offset against foreign income of another controlled foreign subsidiary.
In closing, we again reiterate the importance of the proposed legislation. We must widen the tax base so all taxpayers bear their appropriate share of the tax burden. We should not allow a select group of speakers to ask for a whole list of additional exceptions that essentially undermine what we are attempting to accomplish.
Tax Policy Chief Directorate