We argue that the bill may impact adversely on both vertical and horizontal equity, unless concurrent consideration is given to the impact of provincial taxes on the vertical division of revenue, the horizontal division of revenue between the nine provinces, and administrative capacity issues. Addressing these concerns requires a rethink of the vertical division process and the provincial equitable share formula, which should be changed to incorporate a tax capacity equalisation component, and a mechanism to support those provinces that may not have adequate capacity to administer taxes.

We argue that the bill does not take adequate account of the possible long-term economic impact of provincial taxes. We propose that the additional tax burden of provincial taxation be considered, as well as the contribution of the different spheres and governments to it. This may be facilitated through a regular review process and the publication of relevant data.

We argue that in its current form the bill may not comply with the Constitution, as it fails to regulate provincial taxes as such, and instead seems to imply a one-by-one authorisation process. We propose that the regulatory act should name and regulate specific taxes that provinces can reasonably be expected to impose on the basis of the Constitution, acknowledge the provincial right to initiate any additional taxes that the Constitution does not explicitly prohibit, and provide a review mechanism in instances where the national government has concerns about the economic impact of a specific tax.

Each of these sections is based on longer papers that are appended to this text (appendix A). In addition, appendix B contains nine boxes providing important background facts that may be helpful for Parliament in its consideration of the bill.

1.1. The pending introduction of provincial taxation powers raises at least three sets of equity concerns, relating horizontal equity, the possible asymmetrical implementation of taxes across the provinces, and vertical equity:

1. Horizontal equity. While rich provinces should not be punished, how can one ensure that poor provinces are not disadvantaged due to inadequate tax bases?
2. Asymmetrical implementation. What if only some provinces had the administrative capacity to levy a certain tax, and others not? In this case, provinces that lack the necessary prerequisites might risk forfeiting revenues.
3. Vertical equity. If provinces received substantial powers of taxation, ceteris paribus, they might end up getting in excess of what they require to adequately fulfil their constitutional obligations and functions. Not only would this diminish allocative efficiency; it may also lead to a bias against the national sphere.

1.2. None of these issues are insurmountable. However, they need to be addressed with carefully designed mechanisms:

1. Horizontal equity. An adequately designed equalisation mechanism needs to complement provincial taxation powers. This would require the definition of the basket of tax bases to which equalisation applies, as well as the extent to which capacity should be equalised. A complementary or alternative option could be to strengthen the redistributive impact of the equitable share formula that is used to distribute a part of the nationally collected revenues across the provinces.
2. Asymmetrical implementation. Addressing this issue may involve a set of "tax substitution grants", or a tax point rental system. The national government could give those provinces presently unable to levy a certain tax a grant equal to (or equal to a percentage of) the amount they would have collected if they had been able to impose this tax at an average rate on their given tax base. This should be combined with an incentive element to ensure that provinces develop the prerequisite capacity.
3. Vertical equity. An amount of tax room created by the national government should be set equal to a reduction of the total provincial equitable share, or the two should be linked on the basis of a predetermined mechanism.

1.3. There is a need to take account of the equity environment into which provincial tax powers are introduced. Only then can adequate mechanisms maximise the benefits of provincial taxation, while neutralising the risks. Whether these should be enshrined in a regulatory law on provincial taxes is debatable; but they should at the very least be considered in conjunction with such a law.

2.1. The Constitution recognises the importance of balancing national macroeconomic stability with the power of provinces to introduce taxes. In terms of the government’s macroeconomic strategy, it is necessary to maintain a competitive overall tax to Gross Domestic Product (GDP) ratio, as recognised in the memorandum to the bill.

2.2. The contribution of provincial taxes to the overall tax burden in the short to medium term is likely to be small; a rough calculation shows the potential maximum contribution of the fuel levy and personal income tax surcharges together to approximate 3 per cent of GDP. In addition, the current 25 per cent target ratio seems to be rather arbitrary. It is not clear whether and why an additional 3 per cent would impact negatively on investment or economic growth. However, it would seem prudent for the bill to address a possible situation in the future where an upwardly drifting provincial tax burden pushes up the overall tax to GDP ratio to a point where it might become problematic.

2.3. This raises the question as to whether only the provincial sphere should limit its tax burden. The governance benefits of provincial taxation may be such that provincial taxes are preferable to other forms of income, notably intergovernmental transfers and grants, which would call for an adjustment of nationally collected revenues.

2.4. The creation of tax room will have a knock-on effect on the vertical and horizontal divisions of revenue, which the current budget process is not geared to consider. What is required is a periodic review of the distribution of the overall tax burden between the spheres and the nine provinces. This should be based on intergovernmental consultation, and consider the equity effects of provincial taxation.

2.5. One way to address these concerns in the bill would be through a provision for provincial tax room to be created within an overall specified tax burden, and a requirement for a regular review process.

2.6. In the absence of such measures, policy makers may consider whether (a) the existing budget process could deal adequately with the potential equity issues involved, and (b) whether the bill adequately ensures the availability of the relevant information to Parliament and the public. It is important to note that an evolving system can create a situation whereby it becomes increasingly difficult to obtain accurate information on the distribution of the tax burden, geographically and between population groups. From a transparency and accountability perspective, one would therefore like to see provisions in the bill requiring periodic reporting at the central level on the types and levels of taxes levied by provinces. In addition, a thorough analysis of the contribution of each sphere to the overall level and impact of taxes is desirable.

3.1.The Constitution directly authorises provinces to levy any tax that the former does not explicitly prohibit. The Constitution envisages an enabling Act of Parliament to regulate provincial taxation so that the exercise of these powers does not impede the national government’s macroeconomic management function, economic activity and mobility. In order for this Act of Parliament to fulfil its regulatory function, it needs to address the following three imperatives:

1. Recognise the authority to levy specific taxes. The Constitution already and directly authorises provinces to levy specific taxes.
2. Safeguard the economy. The exercise of these powers may not substantively threaten the overall economy.
3. Allow innovation. The Constitution authorises provinces to levy all taxes that it does not prohibit, and thus enshrines the right to individual initiative.

3.2. The following three provisions could address, respectively, each of the above imperatives:

1. The required Act of Parliament should name specific taxes that provinces may reasonably be expected to levy on the basis of the Constitution, and specify the conditions under which these taxes should be imposed.
2. The bill may fix a maximum tax to GDP ratio, and provide a procedure for determining the distribution of tax burden across the spheres. The concept of "tax room" that the Financial and Fiscal Commission brought into the debate should be further explored in this regard.
3. The bill should provide a negotiating mechanism that could be activated in instances where the national government believes that the implementation of any additional, non-listed provincial tax may "materially and unreasonably" prejudice national economic policy, economic activity or mobility.

3.3. As the bill does not address these imperatives, it does not appear to fulfil the regulatory functions foreseen by the Constitution. This regulation would only occur later by means of what is ambiguously referred to as "regulatory national legislation" (rather than an Act of Parliament) for each individual provincial tax. If this were to take the form of ministerial regulations, it may amount to a usurpation of Parliament’s regulatory obligations outlined by the Constitution. We recommend that, in order to comply with the Constitution, the bill be redrafted to cover the three imperatives spelled out above.

Lindiwe Mbanjwa & Joachim Wehner

Idasa: Budget Information Service

The discussion of provincial taxation powers in South Africa takes place in a context of ongoing and unresolved concern over intergovernmental equity. Developing countries with decentralised governance systems are often characterised by lower spheres of government with limited administrative capacities, and large disparities. South Africa is no exception. Initially, government policy has focused on improving expenditure administration. The equity issues on the expenditure side are daunting enough. For instance, provinces have different capacities when it comes to delivering services, due, in part, to the uneven distribution of the bureaucracies inherited from the former homeland system. Redistribution in favour of poorer provinces has to be balanced with the need to develop sufficient capacity in target provinces to spend scarce resources and the need not to disrupt the ability of richer provinces to deliver services with inter-provincial benefits. Equity concerns are at least as pertinent in the context of the envisaged decentralisation of taxation powers. As an important feature of most rural provinces is that they have augmented delivery needs and limited tax bases at the same time, an insufficient consideration of equity issues on the revenue side may exacerbate current challenges and disparities. The equity challenge requires finding a balance between provinces with tax and delivery capacity and those without. In this context, the following paragraphs attempt to isolate the equity implications that are raised by the implementation of provincial taxation powers, and propose concrete policy options to deal with these issues.

The pending introduction of provincial taxation powers raises, at least, three sets of equity concerns, relating to vertical and horizontal equity, as well as the asymmetrical implementation of different taxes across the provinces. With regard to the issue of vertical equity, if provinces received substantial powers of taxation, ceteris paribus (i.e., if transfers and grants were not adjusted), provinces might end up getting in excess of what they require to adequately fulfil their constitutional obligations and functions. Not only would this diminish allocative efficiency; it may also lead to a bias against national policy. In addition, the national government may not be left with sufficient resources to undertake the necessary and desired extent of income redistribution. On the other hand, if the national government provided tax room, in effect diminishing its own tax income to free up space for provincial taxes, the pool of revenues to be divided between the spheres would be diminished. The continued use of the existing ratios (percentages) for the division of nationally collected revenue between the three spheres would then squeeze the national government, while disproportionately favouring provinces.

A second set of concerns relates to horizontal equity. While rich provinces should not be punished, how can one ensure that poor provinces are not disadvantaged due to the lack of adequate tax bases? The existing equitable share formula has a substantial redistributive effect, although some would argue that its extent is insufficient, given pronounced backlog and need differentials. Simply put, were the equitable share formula maintained in its current form, the effect of the implementation of provincial tax powers would be a strengthening of the overall income of rich provinces, while poor provinces would lose out due to weak tax bases.

A final set of equity concerns is raised by the prospect of asymmetrical implementation. What if only some provinces had the administrative capacity to levy a certain tax, and others not? In this case, provinces that lack the necessary know-how and systems to implement these taxes might forfeit revenues until they have built up the prerequisite capacity (even if a tax base was substantial enough to justify an immediate collection effort). Or should there be a compensatory mechanism? If yes, what form should this take? Because of the lack of choice, this issue differs from a situation where all provinces levy a certain tax, but consciously choose different rates, which would represent an income adjustment at the margin and would not raise equity concerns.

We venture to suggest that none of these issues are insurmountable in the context of revenue decentralisation. However, they need to be addressed with carefully designed mechanisms in order to ameliorate their destructive potential. Pertaining to the issue of vertical equity, the maintenance of the current percentage divisions would favour provinces – they would receive an equal percentage of the pool of centrally collected taxes while gaining access to additional revenues. One option would be that the amount of tax room created be set equal to a reduction of the total provincial share, or that the two be linked on the basis of a predetermined mechanism. This would generate new percentages, and, presumably, more equitable ones – if we assume that the current vertical division is indeed equitable, which some critics would challenge.

There are also well known tools to address horizontal equity concerns, notably by introducing an adequately designed equalisation mechanism. The Financial and Fiscal Commission has long been punting the introduction of a fiscal capacity equalisation grant in conjunction with provincial taxation powers. Many decentralised countries utilise such grants to compensate poorer provinces for the lack of adequate tax bases. What exactly might this look like? One would have to define the basket of tax bases to which equalisation applies, and the extent to which capacity should be equalised. For instance, a basket of the main taxes, applied at average rates, would yield a cross-provincial per capita equalisation standard. Those provinces falling below the standard could be equalised up to this standard amount. In this way, poor provinces would not lose out due to their lack of a tax base, while rich provinces would enjoy access to revenue over and above the average per capita yield. Apart from the introduction of a specific equalisation grant, a complementary or alternative option could be to strengthen the redistributive impact of the equitable share formula that is used to distribute income across the provinces. The most obvious starting point may be to drop the economic activity component from the formula, but further adjustments may well be contemplated. However, it should be noted that rich provinces, according to the Constitution (section 227(2)), may not be "punished" with reduced transfers for own revenue mobilisation. This is in line with international best practice. In Germany, for instance, a recent Constitutional Court judgement ordered a review of the intergovernmental grants system, which has become notorious for its tendency to reward those provinces with poor tax bases, instead of those that contribute to revenue mobilisation. In South Africa, the constitutional requirement will have to be taken into account in any redesign in order to pre-empt similar legal challenges. In addition, there is also a need to produce incentives to encourage poorer provinces to develop their tax bases in the long run, as disparities would otherwise simply be perpetuated.

Finally, the issue of asymmetrical implementation because of administrative capacity differentials needs to be resolved. One solution might be a set of "tax substitution grants", or a tax point rental system. Similar mechanisms have been used, for instance, in Canada (although in a different manner and context). The national government may give those provinces presently unable to levy a certain tax a grant equal to, or equal to a percentage of, the amount they would have collected if they had been able to impose this tax at an average rate on their given tax base. This should be combined with an incentive element to ensure that provinces develop the prerequisite capacity, so that such a system of tax substitution grants or a tax point rental scheme would eventually become redundant after an initial phase-in period.

What this discussion highlights is the need for a careful rethink of the equity environment into which provincial tax powers will be introduced. Only then can we ensure that adequate mechanisms maximise the benefits of provincial taxation, while neutralising the above mentioned risks.

Alta Fölscher

Idasa: Budget Information Service

The potential contribution of provincial taxes to the overall tax burden in the short to medium term is small. However the need remains to provide for a situation in the future where an upwardly drifting provincial tax burden could become problematic at the margin. This would have implications for intergovernmental fiscal relations and the budget process.

The Constitution recognises the importance of balancing national macroeconomic stability with the power of provincial legislatures to introduce provincial taxes. Within a global economy, one aspect of that interest is the need to maintain a competitive overall tax to Gross Domestic Product (GDP) ratio, as recognised in the memorandum to the bill. While national macroeconomic stability is an obvious concern in principle when granting sub-national taxing powers, it is necessary to consider the magnitude of the potential practical impact of provincial taxes in South Africa. A rough calculation involves the two provincial revenue sources with the highest yield potential, the fuel levy surcharge (at the maximum projection of a 40c surcharge) and the surcharge on personal income tax (at the maximum rate of 7 per cent of national income tax). It shows that the maximum take-up by all provinces would add up to about 3 per cent of GDP (on the 2000 GDP).

Whether this is problematic is a matter of opinion. Certainly, the influence it may have as a summary statistic on foreign investment needs to be considered – would a sudden jump to a 28 per cent ratio (if national taxes were at the maximum 25 per cent current target, which they are not) be within internationally acceptable norms? Would this be sustainable in terms of the economy itself? It is however conceivable, especially since local government revenue is not included in the current 25 per cent target, that the slow upward drift of total provincial tax yield at the margin may at some point in the future result in an unsustainable overall tax burden. But this applies in equal measure to local government and even national government. This raises questions on mechanisms in the bill or the existing budget process to manage not only the overall tax burden, but also its distribution between different spheres.

The bill itself is of course a mechanism in this regard. It provides the national government with effective, if indirect, control over the tax bases and rates of any tax. One would assume that the National Treasury would manage the tax in a manner that, if all provinces applied the maximum rate, it would still fall within the desired tax to GDP ratio. However, this may not be a sufficient mechanism should provincial taxation grow significantly. What if the provincial contribution became problematic at the margin and a province brought yet another tax to the table? Would that tax be disallowed? What if the list of taxes in existence favoured a specific block of provinces, as is probable since the more capacitated provinces are likely to get to the table first? And why should the provincial sphere then be penalised? Why could it not have more tax room at that point? The need remains to occasionally review the overall level and distribution of the tax burden as an intergovernmental process.

Such a process would resonate well with the requirement of co-operative governance. It would also deliver a more considered continual balance between national interest and provincial interests, between the imperatives of national fiscal discipline and the economic benefits of decentralised taxation, and between the need to utilise tax bases and achieve regional equity. Besides, it would allow the opportunity to review the overall tax system for progressivity across sphere boundaries. At the very least the level and composition of tax would be the result of considered decision-making and not ex post and incremental. In practice, however, it raises a number of interesting decisions on both the revenue and expenditure side.

On the revenue side: If a tax to GDP ratio was set by the national government it would require tax room to be created for provincial (and local) taxation within that limit – effectively mimicking the vertical split decision on the expenditure side. Subsequently the share of each province within the provincial "envelope" will need to be determined, mimicking the horizontal split decision. Provinces may then have the choice to raise taxes, in any combination from the allowed list, to their assigned limit, as would the national government.

On the expenditure side: Posing no change in the tax or deficit to GDP targets, changes in the relative share of the national and provincial spheres of government in the maximum tax burden could create sudden changes to the available expenditure amounts in the main budget framework (derived from national taxes only and including the provincial equitable share). These changes can either be absorbed in the national budget (spending by national departments) or in the provincial equitable share. If the provincial equitable share decreases, decisions will have to be taken about the distributional effects of the decrease between provinces, articulated with the distributional effects of the provincial own revenue gains. Of course, changes in GDP or tax collection can either ameliorate or exacerbate the expenditure effects of changes in the ratio, which adds fuel to the argument for occasional review.

The design of the review process would require thorough thinking to ensure that its outcomes are in line with constitutional and agreed criteria. For example, where would the final decision as to the vertical and horizontal division of the allowable tax burden rest: with an extended meeting of the national cabinet (i.e., including the provincial premiers)? What would the bases be for taking these decisions – would the horizontal division be based on a formula that allows for ex ante fine-tuning of provincial tax powers in line with equity considerations? What would be the role of Parliament (including the National Council of Provinces)? What mechanisms would be used to make the division of the taxing share binding to some degree? What would the steps be along the way that will ensure maximum opportunity for individual provinces to state their case while allowing consensus building with regard to joint provincial interest and the national interest? What analysis would be required to take well-considered decisions and where would the burden fall to generate that analysis? How much of the information would be in the public domain, and at what stage, to foster healthy public debate and accountability?

What is clear is that the process need not and should not take place on an annual basis. The likely share of provincial taxes would be too small to merit the high transaction costs of a frequent process. Currently fiscal targets are set for the medium term: the intergovernmental tax policy process could take place every three or five years to allow for changes in line with economic and fiscal developments. Principles for the vertical division of revenue and the horizontal equitable share formula can be reviewed simultaneously. Thus the interaction between the distribution of tax and spending can be considered. Besides, it would be an improvement over current arrangements where unpredictable changes to the formula, for example, interfere with the transparency and predictability of funding.

These considerations open up one potential avenue of regulation that is substantive, but not dependent on the nature of individual taxes: requiring provincial taxation to be within an upper limit on the overall tax burden and the design of an intergovernmental process that would emphasise co-operative governance. In the absence of such measures it would be necessary to consider whether (a) the existing budget process and data that supports it can deal adequately with all the potential equity issues involved and (b) whether the bill contains adequate provisions for information availability to Parliament and the public. It is important to note that the evolving system can create a situation whereby it becomes increasingly difficult to obtain accurate information on the distribution of the tax burden, geographically and between population groups. From a transparency and accountability perspective, one would therefore like to see provisions in the bill requiring periodic reporting at a central level on the types and levels of taxes levied by provinces. In addition a thorough analysis of the differential contribution of the spheres to the overall level and impact of taxes will be required.

Joachim Wehner

Idasa: Budget Information Service

Having to regulate the taxation powers of South Africa’s provinces is a challenging task. The 1996 Constitution, according to the highest court in the country, grants provinces specific powers of taxation. But these powers are only implied in the Constitution, which merely lists prohibited sources of provincial revenue. The Constitution allows provinces to levy the residual taxes (i.e., all non-prohibited taxes) within the constraint that the exercise of these powers may not impede the national government’s macroeconomic management function, economic activity and mobility. Furthermore, although the Constitutional Court ruled that the document assigns specific powers of taxation to provinces, it did not clarify the detail of these powers in its certification judgement.

This leaves the national Parliament with a regulatory dilemma. As several legal experts consulted by the Idasa: Budget Information Service confirm, the constitutional obligation that an Act of Parliament is to regulate the exercise of taxation powers by the provinces would, at best, only be partially fulfilled by the Provincial Tax Regulation Bill in its current form. The bill merely outlines a process but fails to regulate the object of the regulatory requirement – the taxation powers as such. At the same time, some argue that provinces can introduce taxes within the parameters of section 228 without the existence of a national regulatory law because of the fact that the Constitution already establishes the provincial authority to do so. The question would then be as to what exactly remains to be regulated in the Act of Parliament that the Constitution calls for.

This article discusses some requirements that regulatory legislation would have to meet, in the author’s view, in order to comply with the demands of section 228(2) of the Constitution. As such, the following should be understood as a suggestion for a compromise between the extremes implied by the laissez faire interpretation on the one hand, and the National Treasury’s desire to retain some level of control over revenue measures on the other. This discussion is preceded by a brief overview of how the regulatory thinking around provincial taxation seems to have evolved in South Africa, and selected thoughts on the impact of regulatory mechanisms on policy outcomes.

The then Department of Finance originally hoped to have a draft framework in place for implementation with the 2001 budgets. The initial intention seems to have been to allow provinces to levy taxation on the basis of an amendable "allowed list". A draft list, which has been circulated within government, includes excise taxes on goods and services, tourism taxes, fuel taxes, environmental taxes, and various smaller items; but it excludes the controversial provincial surcharge on personal income tax.

The bill now before the national Parliament takes one step backwards, and merely outlines a process by which such a list would eventually be arrived at. One possible explanation is the legal uncertainty about the national government’s ability to disallow provincial taxes, notably the mentioned surcharge on personal income tax. This might have led the now National Treasury to propose a vaguer approach that skirts the issue, at least for the moment, by focusing on process requirements rather than specific taxes. Nonetheless, the prior existence of an allowed list indicates that the National Treasury has a clear idea of what taxes it would like to see implemented, and, perhaps more importantly, which ones not. Ultimately, any difference of opinion with regard to the power of the centre to disallow specific taxes that the Constitution assigns to provinces may have to be settled in court.

Regulatory framework legislation can significantly shape policy outcomes, and the balance of power between organs of state or levels of government. The extent of and manner in which this occurs is determined by the choice of concrete regulatory tools. Notably, rather than providing substantive guidance, political actors can control administrative outcomes through process. Mechanisms that can stack the deck in order to impede policy innovation include procedures requiring substantial resources in order to be complied with. The more cumbersome such procedural hurdles, the less likely is policy change.

In the context of provincial taxation, elaborate procedures that provinces have to follow will delay the implementation of provincial taxation powers, and stretch the maintenance of the status quo (i.e., the absence of provincial taxation). By putting the "burden of proof" on provinces, the draft bill uses deck stacking against the provincial sphere. The reverse effect would be achieved, for instance, by imposing the procedural burden of proof requirements on the centre in (selected) cases where it believes that its macroeconomic management function is threatened. In the latter case, the deck would be stacked against the centre and in favour of provinces, by placing the administrative obstacle course in front of the national government.

How can one assess the adequacy of the draft bill on the basis of the Constitution? The following are key principles that have to be recognised, in the author’s view, in order for an Act of Parliament to constitute the enabling legislation envisaged under section 228 of the Constitution:

• Recognising specific tax authority. The Constitution already and directly authorises provinces to levy specific taxes, although it does not name these (section 228(1)).

• Safeguarding the economy. The Constitution places a responsibility on the national Parliament to ensure that the exercise of provincial taxation powers does not substantively threaten the overall economy (section 228(2)).

• Allowing innovation. The Constitution authorises provinces to levy all taxes that it does not explicitly prohibit, and thus enshrines the right to individual initiative. The creative use of this constitutional space opens up a (theoretically endless) multitude of possible tax options.

The draft bill only addresses the third principle, as it does not limit the potential for provinces to propose certain taxes. However, the bill does not address the fact that the Constitution directly grants provinces specific powers of taxation, nor does it seem to satisfy the constitutional requirement that the tax powers as such are to be regulated in order to safeguard the economy. Under the draft bill, the latter would only occur by means of what is dubiously referred to as "national legislation" for each individual provincial tax. If this were to take the form of ministerial regulations, it would most likely equate an unconstitutional usurpation of Parliament’s obligations under section 228(2)(b) of the Constitution.

Given that the draft bill seems not to meet at least two of the regulatory requirements, exactly what would have to be contained in an adequate law?

• In order to recognise the constitutional authority of provinces to levy specific taxes, the required Act of Parliament should name specific taxes that provinces may reasonably be expected to levy on the basis of the Constitution (and not on the basis of the political preferences of the centre). The regulation of the exercise of provincial taxation powers would mean that the act has to specify the conditions under which the specific taxes already in the legislation should be imposed by individual provinces, by outlining tax base, rate bands, exemption possibilities and procedures, etc.

• The bill may fix a maximum tax to GDP ratio, and provide a procedure for negotiating the distribution of tax burden across the spheres. In this way, Parliament would honour its constitutional obligation to maintain macroeconomic integrity and cohesion during the exercise of provincial taxation powers. The concept of "tax room" that the Financial and Fiscal Commission brought into the debate should be further explored in this regard.

• The constitutional right to tax creativity might be recognised by providing a procedure to deal with requested variations of those taxes already specified in the legislation, and to deal with suggestions for additional taxes not yet specified in the legislation. The draft bill contains a procedure not unlike one that might be used to meet this requirement. On the other hand, it is possible to argue that even the inclusion of such a procedure may contravene the right to individual initiative granted under the Constitution, if it would seem to lead to an individual "authorisation" of specific taxes. An alternative option may be to clearly state that provinces may deviate from the guidelines provided for specific taxes, as well as introduce any other taxes not explicitly prohibited by the Constitution. This scenario may leave it to the bill to provide a negotiating mechanism that could be activated in instances where the national government believes a provincial tax may "materially and unreasonably" prejudice national economic policy, economic activity or mobility, which are the constraints imposed by section 228(2)(a).

Despite several options that remain open in the above discussion, we can assume that a bill along these lines would substantially redistribute the burden of proof, compared with the current draft. This seems more in line with the direct constitutional authorisation of specific provincial taxes, which no process or substantive legislation may disregard.

I have argued that legal uncertainties and political preferences led the National Treasury to propose a peculiar, process-focused approach in attempting to fulfil the regulatory obligations with regard to provincial taxation. Furthermore, the particular process requirements that are suggested in the draft bill, compared with alternative options of regulating processes, are likely to lead to policy outcomes that reflect the interests of the centre.

In conclusion, I would argue that the current draft fails, at least for the larger part, to regulate the provincial power to tax. If passed in its current form, this could lead to protracted legal wrangles and uncertainty. Taken together with the implied constitutional requirement that the enabling legislation be passed within a reasonable time period, a sustained regulatory vacuum due to court challenges might give rise to an uncoordinated implementation of provincial taxation, based on the disparate initiatives of individual provinces. Some may consider this to be undesirable. An adequate regulatory effort is called for as a matter of urgency.

In the United Kingdom, the Scotland Act (1998) initiated the devolution of substantial expenditure responsibilities to the Scottish Parliament. Apart from receiving a block grant from the central government, the Scottish Parliament has powers to vary the rate of personal income tax for Scottish residents by 3 per cent (based on the national personal income tax). It is estimated that a 1 per cent change in the basic rate is equivalent to about £150 million in tax revenue at present. In the debate on the White Paper on a Scottish Parliament, the then Secretary of State for Scotland, Donald Dewar, explained the central government’s expectation: "It is important to recognise that the power may be used to deal with some special project or difficulty. I do not expect that it would simply be added to the block-and-formula sum that is negotiated as a continuously available additional revenue support. I believe that that would constitute a misuse of the power - although, of course, it would be a matter for the Scottish Parliament. The power must be seen in the context of the United Kingdom direct taxation system as a whole."

Source: House of Commons Debates, 31 July 1997, col. 465.

Provincial taxes

228. (1) A provincial legislature may impose ­

1. taxes, levies and duties other than income tax, value-added tax, general sales tax, rates on property or customs duties; and
2. flat-rate surcharges on the tax bases of any tax, levy or duty that is imposed by national legislation, other than the tax bases of corporate income tax, value-added tax, rates on property or customs duties.

(2) The power of a provincial legislature to impose taxes, levies, duties and surcharges ­

3. may not be exercised in a way that materially and unreasonably prejudices national economic policies, economic activities across provincial boundaries, or the national mobility of goods, services, capital or labour; and
4. must be regulated in terms of an Act of Parliament, which may be enacted only after any recommendations of the Financial and Fiscal Commission have been considered.

Source: The Constitution of the Republic of South Africa (Act 108 of 1996), section 228.

"Without detracting from the important efforts SARS [the South African Revenue Service] is currently undertaking in its restructuring with a view, inter alia, to the enhancement of its capacity, it is correct to state that at present SARS is still experiencing problems with regard to its capacity. Accordingly, for the same reason that other developing countries have not burdened their tax administrations by the imposition of a surcharge on the PIT [personal income tax] base, the Commission has serious concerns about doing so in South Africa. To do so in the short term could give rise to a risk… A possible alternative to the surcharge on the PIT base is a provincial surcharge on the national fuel levy. This levy could be raised with relative ease at a lower-tier level, provided that retail outlets have pumps with sealed meters that can be monitored effectively… The obvious mobility of motor vehicles will be a check on excessive surcharges. Nevertheless, this option would allow lower-tier governments greater discretion over their revenues than the present system."

Source: Commission of Inquiry into certain Aspects of the Tax Structure of South Africa: Seventh Interim Report, Synthesis of Policy Recommendations with regard to Provincial Taxation (July 1998), paras. 2.9 and 3.1.

"[439] It is apparent that the national legislation envisaged under NT [New Text, i.e., the 1996 Constitution] 228(2) is to ensure the coherence of the taxing system and is not directed at providing the underpinning of the taxing power itself… The term "authorised" is used to signal the empowerment by law or the courts. "Regulation" however, is habitually used in statutes in conjunction with the word "control" to signify the object of legislative authorisation, the directing and commanding of that which has been authorised to be regulated. Thus seen, NT 228 affords provincial legislatures specific and guaranteed taxing powers. The IC [Interim Constitution] offers provinces merely the expectation of such powers. It is by reason of the greater specification and detail in the NT that certain types of taxes, levies and duties have been omitted from provincial legislative competence, but this omission is more than offset by the assurance of specific taxing powers…"

Source: Constitutional Court of South Africa, case CCT 23/96: Certification of the Constitution of the Republic of South Africa, 1996 (6 September 1996), para. 439.

• The bill distinguishes provincial taxes from user charges. A provincial tax is defined as a compulsory payment to a provincial government imposed in terms of section 228 of the Constitution, excluding a user charge. A user charge is defined as an identifiable payment for a specific good or service provided to a willing buyer that directly links benefit and fee to the person making the payment.
• A provincial tax must comply with the provisions on provincial taxation and the principles of co-operative governance in chapter 3 of the constitution.
• A province intending to impose a provincial tax must submit particulars of the proposed tax to the Minister of Finance at least ten months before the start of the next fiscal year. These must specify the tax base and desired rates, the tax collecting authority, revenue projections on a quarterly basis over three fiscal years, and economic impact assessments, amongst others.
• The Budget Council and the Financial and Fiscal Commission have to be consulted.
• If a tax proposed by a province complies with all requirements, the Minister of Finance tables regulatory national legislation for a provincial tax, which prescribes the tax base, rate band, and collecting agent. The legislation is to enable all provinces to impose the tax, but any province may choose not to do so.
• The South African Revenue Service (SARS) is the default collection agency for provincial taxes. A province may not impose a tax unless it has concluded an agreement with SARS or an alternative collection agent.
• The Minister of Finance is not required to complete his review of a provincial tax proposal within a specific time period.

Source: Provincial Tax Regulation Bill, 2001; Government Gazette No. 22353, 1 June 2001.

• "The Commission reaffirms its previous recommendation, based upon the current constitutional dispensation, that the following taxes constitute the most feasible form of provincial own revenue sources: a surcharge on personal income tax, a fuel levy, and betting and gaming taxes. The surcharge on personal income tax is the most significant of these revenue sources. Decisions on the provincial tax structure should take account of revenue assignment in the local government sphere.
• Provinces should be allowed the flexibility to determine their own tax rates within the bands determined by the Minister of Finance.
• To implement any of the provincial own revenues measures, tax room must be created in order to maintain the tax burden within nationally determined targets."

Source: Financial and Fiscal Commission: Submission on the Division of Revenue 2002, 2003 (July 2001), p. 23.

A surcharge on the national personal income tax;

taxes on private sector health institutions, health care providers and dentists;

a tax on the disposal of hazardous material;

a development levy;

a tax for heritage protection;

taxes on alcohol and tobacco;

taxes on heavy road freight transport or motor vehicles;

a turnover tax;

a levy on agricultural goods;

a tourism bed levy;

a tax on businesses that emit fumes;

a tax on alien vegetation;

a landmark levy.

Source: Business Day, 17 January 2001.

Source: Adapted from Department of Finance, Canada: http://www.fin.gc.ca/FEDPROV/feqe.html