SAJEMS NS Vol 4 (2001) No I The Political Economy of Capital Gains Taxation in South Africa I Part I: The Public Finance of Capital Gains Taxation Zane A Spindler Simon Fraser University. Canada Visiting Professor. School of Economics. University of Cape Town 2001 ABSTRACT Public Finance and Public Choice principles are used to analyze the ideological and practical basis for the proposed introduction of a Capital Gains Tax into the income tax system of South Africa. The paper concludes that this is a flawed tax whose time has passed - especially for countries like South Africa. JELH24 A foolish consistency is the hobgoblin of little minds adored by little statesmen and philosophers and divines. With consistency a great soul has simply nothing to do. RW Emerson, Spiritual Laws, 1803-82. 1 INTRODUCTION In its 2000 Budget, the South African national government proposed that a capital gains tax (hereafter, CGT) be introduced as of I April 2001. In its 2001 Budget, the Minister of Finance said: "While it is imperative that this fundamental change be made, it is proposed that the implementation be deferred to I October 2001." With luck, and some greater appreciation of post-modem public finance, by the time I October 200 I arrives, the CGT proposal will be shelved permanently. When the current proposal was first introduced, the usual rationales were given for adding such a tax to an income-based, ability-ta-pay, tax system; namely, the improvement of consistency and equity of the tax system. This essay will analyze those motives to suggest that they are based on misconceptions of consistency and equity; namely, 1) that achieving consistency on one margin necessarily involves creating inconsistencies on other, potentially more important, margins, and 2) that achieving equity should be judged with tax R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . 2 SAJEMS NS Vol 4 (2001) No 1 incidence in mind. These propositions are based on the well-known economic theories of the second best and tax incidence. Further, the essay will argue that the supposed problems actually arise from existing features of the income tax system; namely, I) the lack of full integration of the personal and corporate income taxes and 2) the difference between the average and marginal tax rates. Changing these features directly may be more appropriate as a way of establishing consistency and equity, as well as possibly raising more revenue (Hall & Rabuska, 1995). In addition, consideration will be given to the political basis for introducing and maintaining a capital gains tax as part of a tax system. Initial support may quickly erode as the incidence of the tax and its real burden becomes more extensive and more apparent. Experience from other countries with capital gains taxation suggests that, in general, net revenue prospects may not justify the introduction of what amounts to a tax on capital rather than on capital income (Grubel, 2000; Katz, 1996). That experience also suggests that capital gains taxation often evolves into a more intrusive and distorting fonn over time. Finally, the essay will argue that globalization of the capital market has made the taxation of capital by developing countries a policy which is most beneficial to the most developed countries that already have such taxation in place, rather than the developing countries that are attempting to hannonize their tax systems as a conduit to "joining the civilized world". This essay will explore many novel ideas not usually brought into play for tax policy analysis. The latter generally pays more heed to standard public fmance and accounting traditions. These alternative perspectives suggest that the capital gains tax proposed for South Africa (hereafter, ZA) is the wrong tax at the wrong time for the wrong country. The essay will be organized and published in two parts, of which this is Part I. Part I has two major sections beyond the Introduction and ends with a·· conclusion. These next two sections will look at taxation, in general, and at CGT,in particular, from a modem Public Finance perspective. However, they will take an unorthodox twist by considering the internal and joint consistency, and the appropriateness, of public finance principles used to justify the inclusion of a CGT in actual taxation systems. In Part II, there are two more major sections. One of these will consider Public Choice perspectives on tax policy decisions with respect to including a CGT as part of a polity's tax system. The other will discuss broader perspectives of the CGT for South Africa's global future and macroeconomic growth prospects. R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . SAJEMS NS Vol 4 (2001) No 1 3 These will take what I choose to call a post-modern, political economy perspective. A brief summary and concluding remarks will end Part II and the essay. What follows in both Parts will tend to be more of a philosophical nature about capital gains taxation, in general and for countries like South Africa in particular, rather than a set of specific criticisms of the current South African CGT proposal (Such specific critiques can be found in Black:, 2000; Grote & Fletcher, 2000; and Steenekarnp, 2000). The South African capital gains tax proposal (hereafter, ZA-CGT), in any case, may well change substantially before or after implementation. The force of most of the arguments in this essay will not be blunted by any likely modifications, other than its permanent shelving, of course. However, references to specific features of the ZA-CGT are included where the general principles apply. 2 THE PUBLIC FINANCE OF INCOME TAXATION As a discipline, Public Finance simultaneously offers a number of principles or ideals of taxation that are mutually inconsistent, not only on the basis of logic, but also on the basis of its own analysis. Once these inconsistencies are noticed, one is tempted to suspect that, if public finance specialists are not schizophrenic, they at least must have what psychologists call logic-tight compartments. Otherwise, it seems they might have been more hesitant about using these principles jointly as a rationale for specific taxes or tax systems. Of course, it may simply have been a matter of not challenging a received discipline (Who are you to disagree?); conventionalism (That is the norm, follow itl); division of labour specialization (Hiding the forest from the trees.); or worse, not undermining professional market potential for giving expert advice (Successful experts must not appear ambivalent!). As appealing as these spurious arguments may seem to the opponents of CGT, this essay will offer a Public Choice rationale that suggests individual rationality or consistency, but systemic irrationality and inconsistency. Here is the basic consistency problem as it applies to income taxation in general and CGT in specific (Chant, 1999). Most simply stated, it is that the accepted ideal about what to tax is not compatible with the accepted ideal about how to tax. The explanation is a bit more complex. First, we will consider the "what to tax" issue and its own inconsistencies. Then we will explore the "how to tax" issue. R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . 4 SAJEMS NS Vol 4 (2001) No 1 Without dwelling too much on its own rather shaky foundations, we will show why it cannot be achieved given the choice of what to tax. 1.1 What To Tax Since at least the publication of R A Musgrave's modem classic The Theory of Public Finance text in 1959, which really defmed Modem Public Finance, mainstream public finance theorists have considered the appropriate base for income taxation to be what is known as the "Schanz-Haig-Simons" (hereafter SBS) notion of comprehensive personal income (Katz, 1996; Haig, 1921; Simons, 1938). As South Africa's own Katz Commission noted: under this defmition, personal income "is given by the market value of consumption plus the net change in the value of assets or property rights during the period in question" (Katz 1996). Since capital gains either increase the net value of an individual's assets, or allow a higher level of consumption, the SBS comprehensive defmition implies that capital gains should be regarded as income and taxed as such. In 1967, this accrual definition was given practical importance by Canada's Carter Commission, which advocated its application to the Canadian income taxation system, and popularized the comprehensive definition under the tautological slogan "A buck is a buck". Since then, the SBS comprehensive, consumption plus net accretion, definition has featured prominently in almost every income tax reform proposal - especially for those concerning capital gains. Indeed, the Katz Commission harks back to these sixties' notions to provide the ideological rationale for CGT and other proposals. The underlying idea is simple. People should be compared on the basis of their potential well-being during a given tax year. A person's potential well-being is thought to depend on the maximum consumption she could undertake without changing the value of her assets. If saving is assumed to be zero, maximum potential consumption would then equal that year's income, measured on an accrual basis. This seems like a very good idea as long as one does not think too hard about how this would apply in the real world. That is because SBS is based on the simplest possible, and most easily understood, economic model of human behaviour - the one-period model. More realistic two-period, life cycle, or overlapping generation models would not generally consider a given period's (year's) consumption and net wealth accretion to necessarily represent income earned during that period. Indeed, professional and academic economists generally believe that an asset's market R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . SAmMS NS Vol 4 (2001) No I 5 value represents an associated discounted expectedfuture income stream. which is not necessarily related to its associated income flow, or any other income flow, during the past tax year (Lang & Shackelford, 2000). The SHS concept of yearly taxable income may well have seemed more appropriate when it was first developed at an early stage of economic modeling (Smetters 1999) - especially given the requirements of the fiscal budget cycle for yearly revenue. However, it is hardly appropriate as a theoretical concept today, when multi-period modeling is so common both theoretically and practically, and Second-Best Theory suggests that improving consistency on some but not all margins is not necessarily desirable. Further, SHS has never been practical for direct, consistent application. SHS's accrual notion of income is seldom applied because of the impracticality of accurately measuring asset values yearly and of taxing incremental value. In practice, most income taxation is on a realized, not an accrual basis, and unusual gains or losses are often averaged over previous years or carried over to subsequent years. Further, it turns out that "a buck is not often fully a buck" when it comes to treating certain realized receipts as income. Canada, for example, counts a receipt of employment income, capital gains, and lottery winnings as 100%, 75%, and 0% per cent of "bucks", respectively, for pwposes of detennining taxable income. So much for the Canadian legacy of the Carter Commission! Treating such obvious windfall gains as those from lotteries, on a more favourable basis than the treatment of real capital gains due to successful investment, or even than phantom capital gains due to changes in absolute or relative prices, is hardly consistent with the comprehensive definition of income. Unfortunately, the South African Government's CGT (hereafter ZA CGT) proposal follows this very same inconsistent treatment - perhaps with an eye to the revenue generating potential of its new lottery and the lottery's interaction with the tax system. 2.1.1 An Instructive Example: The SHS notion of comparing individuals' yearly income on an accretion basis does not successfully meet the theoretical desiderata of treating equals equal1y. This can easily be seen by the comparison of a production worker who, say, receives employment income of R40 000 per year, with a retiree living off the interest on bonds of R20 000 per year. Suppose during the current year the retiree experiences a R20 000 capital gain on the market value of her bonds due to a downward deviation from average of the interest rate during the tax year in question. Then both the retiree and the R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . 6 SAJEMS NS Vol 4 (2001) No 1 If the retiree did not have to worry about maintaining her consumption during following years, she could indeed consume as much as the production worker during this specific tax year by selling off a portion of bonds to realise her momentary capital gain. However, that would require the retiree to be very short-sighted (or short-lived). Further, income tax systems supposedly based on the comprehensive, accretion concept, recognize the harm and burden imposed if the concept were strictly applied in this case; they do not in practice treat the capital gain on bonds as income until it is realized. 2.1.2 A Common Mistake: Now a casual observer might be inclined to think that the retiree was somehow better off than the worker, because she has wealth in the form of financial assets, which the worker does not. Professional and academic economists - especially those in public finance - should not make such an elementary, wealth-measurement mistake. De facto, the worker owns his own human capital, the value of which would be measured by capitalizing his R40 000 at the appropriate interest rate. This latter interest rate may be higher than that used to capitalize bond coupon payments given that human capital is not alienable in the same way, and, hence, is not as liquid, as fmancial capital. However, the capitalization rate would have to be double the rate applicable on bonds for the value of the worker's wealth to be less than the retiree's wealth. Further, a general decline in interest rates raises the capital value of both human capital and financial capital. Both the worker and the retiree will experience an increase in wealth and both could potentially increase their consumption during the period in question. Both the retiree and worker could borrow more against their respective assets, albeit at different interest rates. 2.1.3 Please Note: The important point to consider about this example is that both individuals are capitalists who have experienced a temporary capital gain during the period in question. Therefore, including the unmeasured worker's capital gain on the same basis as including the retiree's unrealized capital gain would make them distinctly unequal for tax purposes. The retiree has a lower taxable capacity than the worker, which would be recognized by taxing only realized money income - as would be done by an income tax without CGT. Fairness under the comprehensive accretion concept would require including the worker's unmeasured and imputed capital gain on the same basis as the retiree's unrealized capital gain. However, it would be far simpler, and equally fair, to forget about this worrisome comprehensive definition, and treat both on a realized income basis. This is exactly what the ZA-CGT proposal does, despite its SHS rationalization! R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . SAJEMS NS Vol 4 (2001) No 1 7 Thus, real world, income-tax systems, supposedly based on the comprehensive accretion concept, quite sensibly ignore both changes in period-spending potential. However, such tax systems do treat realizations differently as a result of including the realized capital gains from marketable capital, but not the realized in-period increase in borrowing and purchasing power associated with temporary interest rate reductions. Indeed, some macroeconomic and real business cycle models imply that workers' labor supply will vary directly with interest rates so that this worker's labor income may actually be lower than normal or permanent income in this situation. Hence, applying the comprehensive, accretion concept on a yearly rather than life cycle basis would promote a vertical inequity as well as a horizontal inequity, if judged from the longer perspective. Finally, if interpersonal equity is a goal of a tax system, rather than inter-income equity, the SHS comprehensive definition must be applied over individual life cycles or even over generations. Indeed, the more recent research on income and capital gains taxation does consider intergenerational modeling and, in one recent article (Atkinson et al., 1999) finds that the optimum rate for a CGT is zero! Also, intergeneration equity has been attacked from a slightly different angle, (possibly unintentionally) designed to provide the basis for inter- generation rent seeking, by Corak (ed.) (1998). If the supposedly common pattern of "rags-to-riches-to-rags in three generations" generally applies, in-period redistribution would be less appropriate. Real life is less like a hundred-metre dash, than like a relay race, or even more like a series of dashes and relay races. Sensitive handicapping in taxation, as in sport, would work better if it takes account of such realities. We do not demand that, at each change of the baton in a relay race, the differences so far achieved be reversed. That would change the nature of the race and relay- race fans would think it quite unfair! In Life's race to the top, we should also be wary that inappropriate interventions reduce the incentive to race. From that perspective, we would see that progressive tax rates and capital gains taxation, in practice, have more to do with holding back the poor than with pulling down the rich, whose relative position is generally unchanged as a result. We would benefit more, in general, without such anti-productive, anti-growth impediments. Somehow the wrong, static, conception of the world has crept into, and taken up residence in, modem policy analysis, just like the proverbial camel creeping, small end first into the tent. For the sake of the poor, and the tax-system disadvantaged, this textbook notion must be purged and replaced with a realistic, dynamic, conception. R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . 8 SAJEMS NS Vol 4 (2001) No 1 In summary, the SUS, comprehensive-accretion definition of yearly taxable income does not have a sound foundation in economic and public fmance theory and it is impractical and often ignored in practice. In fact, its sole practical purpose seems to have been to provide an ideological rationale for including capital gains in income tax bases. However, a principle, which is so inconsistent and discriminatory in application, is really no principle at all. 2.2 How To Tax There are two classic principles of taxation which modem public fmance has vested with special meaning: The benefit principle and the ability-to-pay principle. The benefit principle (hereafter, BP) states that taxes should be paid by those whom benefit from government provision of goods. In the case of government- provided, private goods, the taxes should simply be user fees similar to the market prices that would prevail if such private goods were privately provided. In the case of what have come to be known as pure-public goods - goods which are non-exhaustive and non-excludable - taxes would also be like user fees but not directly comparable to market prices which either may not exist or may not determine the relevant margin for such goods. Because of the alleged non-excludable and/or non-exhaustive character of public goods, it is presumed not possible and/or desirable to collect prices for such goods, once they are provided, because of the so-called free-rider problem. Supposedly, those who benefit from a good may enjoy it whether or not they have paid for it. Thus, there is an incentive not to payor to free ride. Since private, for-profit firms require payment to survive and prosper, there may be too little, or no, incentive for them to provide public goods. Consequently, this free-rider hypothesis provides the ideological rationale for government, and government provision of public goods, fmanced by taxes.2 In its modem fonn, the BP requires that such taxation be according to individual marginal benefit Given the free-rider problem, individual marginal benefit may not be readily apparent, but there are procedures to fmd an approximation to the BP in practice (Gradstein, 1999; Rondeau et al., 1999; Hines, 2000; Knight, 2000). With these procedures for pUblicly-provided, public goods and with user fee pricing for pUblicly-provided private goods, free riding is controlled, BP applied, and tax avoidance is avoided. However, these proposals have seldom played a major role in modem tax systems or in tax reform proposals. This is probably because they are ill suited for financing a government's redistribution activity, especially when redistribution has little or no efficiency rationale, as in the case of pure transfer seeking (Tullock, 1981). R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . SAJEMS NS Vol 4 (2001) No 1 9 Enter the second classic principle of taxation. Originally, the Ability-To-Pay (hereafter, ATP) principle was probably based on, or evolved from, pragmatism or administrative economy. A money tax could simply not be collected where there was no money. However, ATP need not only refer to taxes in money, but also in kind, as was more common historically before the emergence of money-based economies. Equal sacrifice in kind for (say) labor time might lead to equal taxes on all individuals since nearly every individual at birth has an equal expected endowment of "person-years of labor time". Or, equal sacrifice in terms of burden or disutility of labor might tax the rich less because their time was more valuable in other pursuits! Not so strangely, these alternative, equally valid, concepts of ATP have not succeeded politically in modem times. Instead, modern public finance attempted to give ATP more popular appeal, and moral justification, by relating ability to utility and taxation to sacrifice or burden as measured by utility forgone. Horizontal and vertical equity concepts were interpreted in terms of equal sacrifice of utility or equal burden in terms of forgone utility. Whether utility sacrificed should be equal total, equal average (proportional) or equal marginal has never been (nor could it be) determined in abstract without political considerations. However, none of these measures is operational under economically reasonable and politically acceptable assumptions. Yet, egalitarians were most pleased with the equal-sacrifice reformulation because, under certain assumptions about the shape and commonality of utility functions, disproportionate taxation of the rich, indeed, even a levelling of incomes, could be "scientifically" justified. The fact that the enabling assumptions could not generally be justified, much less measured or empirically validated, typically is mentioned satta voce and, thus, gave faint pause to "the emperor's fashion consultants". Thus, progressive tax rates are the norm for income tax systems throughout the world and generally justified on some notion that an extra dollar is worth less to a rich person than to a poor person. Not so strangely, the buzz-phrase, "a dollar is a dollar", is not applied when the dollar in question is held by people of different means. However, even here there is not consistency. Public finance specialists sometimes argue that credits should be used instead of deductions because deductions would deliver more dollars to those taxed at higher, progressive rates as indeed they should. If comparable diminishing marginal utility determines R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . 10 SAJEMS NS Vol 4 (2001) No 1 both the rate structure and that each dollar is worth less to the richer person, then more dollars in compensation are required for those richer in order to yield an equal sacrifice to those poorer. That is, consistency would dictate deductions not credits. Political and financial expediency dictates otherwise. More importantly, ATP is not applied at the point of final burden and this is the crux of the fundamental inconsistency between the SHS concept of taxable income and the ATP principle of taxation, as both are currently applied to justify existing income tax structures or reform proposals. The SHS concept concerns where the income tax is levied, while the sacrifice or burden concept of A TP must be judged by where the tax ends up creating a burden or sacrifice. However, modem public finance most scientifically valid theoretical and empirical exercise, tax-incidence analysis, generally concludes that these are not at the same place for many types of income - especially capital income. That is, it is inappropriate to judge equity on the basis of who initially pays the tax. 3 THE PUBLIC FINANCE OF CAPITAL GAINS TAXATION Simply stated, the problem is as follows for the capital gains tax. A fully anticipated tax placed on owners of capital, in a small country that must compete internationally for capital, in the long run will not be borne by the owner of capital, but rather by the workers and consumers of that country (see Graham, 1999; Hindriks, 1999; Jones, 2000; Kneller, 1999). Thus, an attempt to use capital taxation to reduce the after-tax income disparity between the rich and the poor may have exactly the opposite effect. The before-tax income of the working poor will fall - possibly as much or more than the after-tax income of the rich who still choose to remain in, and/or migrate to, this country. If the tax is not anticipated, those who already have their capital resident in the country will suffer a capital loss as the new tax is capitalized. But creative accounting, gain-loss realization timing, and official corruption can minimize that burden or even shift it back on the tax system and the general taxpayer. The same result would apply with respect to new taxes, or taxes at higher progressive rates, applied to the incomes of owners of specialized human capital assets which are internationally marketable and internationally mobile. These human-capital owners would immigrate, or fail to emigrate, to the point where local supply was reduced enough to raise their before-tax incomes and restore their after-tax compensation to previous levels. An important point must be made here, given recent misleading public statements by some Ministry of the Treasury (formerly Department of Finance) R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . SAJEMS NS Vol 4 (2001) No 1 11 and South African Revenue Service officials. Just because some foreign countries already have a CGT, even if at higher tax rates, does not mean that foreign and domestic investment in South Africa will not be affected by South Africa's adoption of a CGT. Any major event that changes the expected risk- return balance of investors' portfolios will eventually provoke a portfolio adjustment. Such adjustments may well involve increased net capital outflows, the magnitudes of which are difficult if not impossible to measure and predict accurately. Even in a closed economy, there would be a similar reduction in the well-being of the working poor, as well as for the economy in general. Both rich and poor individuals would have less of an incentive, and ability, to save and invest in either human or non-human capital that improved the formal sector market productivity of the economy. There would instead be substitution into leisure and non-market activities and investment. These very well-known and thoroughly studied incidence results3 from modern public finance are seldom juxtaposed against the similarly well-known, studied, and advocated SHS and ATP concepts. As my colleague, John Chant (1999), has wittily observed, public finance specialists "fail to practice what they teach!" As in other areas of economics recently, there may be some redistribution "policy irrelevance" theorems still to be explored, and claimed, in post-modern public finance. Hopefully, this discussion has made it clearer that: The SHS comprehensive concept of income is not theoretically or practically useful in guiding taxation aimed at establishing either horizontal or vertical equity with respect to real persons. Further, SHS has been most misleading about the very nature of capital gains. It is time for greater clarity. 3.1 What [s Income? Economists will generally say that income is a flow of services from a stock of assets, such as human and non-human capital, as embedded in humans, intangible assets (bonds, stocks), and real assets (including land and buildings and equipment thereon). These real service flows may be given nominal values by the use of their respective, nominal market prices or may be assigned a real value, individually or in total, if appropriately weighted by their relative (or real) market prices. R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . 12 SAJEMS NS Vol 4 (2001) No 1 National accounts economists would further stress the distinction that these service flows contribute to value-added to the newly produced goods and services in the circular income and expenditure flow. That is, from a national accounts perspective, an individual's yearly income is the sum of her value- added to measured Gross Domestic Product (GDP). The economy's net income is similarly the sum of yearly net value-added to Net Domestic Product (NDP). For those who seek consistency in taxation, it is curious that the value-added standard has not played a more prominent role. It would be consistent with national income accounting standards. It would also make empirical macroeconomists very pleased to be able to reconcile government fiscal action with the national accounts' effects thereof. However, this standard is not without its own problems for actual taxation, since it is also accrual based. But many of these accrual problems, like that of depreciation or imputed income, are already dealt with creatively by both national revenue and national income accountants - the latter usually without resorting to counting capital gains. Specifically co-ordinating these efforts and standards would go a long way toward achieving consistency between two branches of government the national revenue and national statistics branches. We must note before leaving this section, capital gains are not included as income in the national accounts, for the very good national accounting reason that: Capital gains are not income. Thus, a value-added concept of income may be more practical than a SUS comprehensive concept, and it would not require or justifY capital gains taxation. 3.2 What Are Capital Gains? Simply stated, capital gains (losses) are the increase (decrease) in the market (or implicit) value of an asset, or set of assets, between two dates. For an asset whose intrinsic nature has been neither augmented nor diminished between such dates, there has been no real capital gain. The "asset is the asset", just like "the dollar is the dollar". 3.2.1 Temporary Gains: The change in the asset's market value may be only due to temporary and reversible changes in relative prices between assets, relative prices between assets and consumption flows, and relative prices between present values and future flows (more commonly called interest or discount rates). R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . SAJEMS NS Vol 4 (2001) No 1 13 Reversible, temporary changes in value do not represent appropriate tax targets if full loss offset is allowed. Revenue gained on the upside of the cycle is lost on the downside, with a great deal of private and public administrative cost in between. One might take exception with this statement, in the case of traders who make their current income from trading in such markets. To the extent that they are successful, their income flow will always be positive. However, their net gains are balanced by others' net losses for a perfectly symmetrical cycle. Partial loss offset may allow some net revenue, but only by playing the intrinsically unfair game "Heads, I win, Tails you lose" with the taxpayer. That game is never popular and often induces extra lobbying and political costs, evasion, and disrespect for law and government, when losses are concentrated for any group. Again, unfortunately, that is the "game" that the Minister of Finance proposes to "play", in part, since losses under his ZA-CGT proposal will only be allowed to offset capital gains, with a carry forward, but an offset against other income is not allowed. 3.2.2 Permanent Gains: Permanent, or long-term, changes in relative prices may require changes in asset holders' portfolio composition or their production or consumption mix in order to maintain efficient allocation when there are variable rather than fixed proportions that apply in asset holders' portfolio balance, production, and/or consumption functions. This may require realization of supposed capital gains, but will not necessarily represent a welfare gain for asset holders. Indeed, the readjustment may be required just to maintain efficiency in the face of otherwise larger welfare losses. The information required to make a scientifically based judgement here is usually too costly, if not impossible, to obtain. But otherwise, the presumption of welfare gain is not warranted. Perhaps some examples here would clarify what might otherwise be an obscure point to non-economists not familiar with isocost, isoquant, and isoutility curves and the Hicksian concept of real income. Example I: First, let us take our retiree, who (say) bought a Sea Point flat in the days when Sea Point was a much different place than it is today. The initial intrinsic or real value of the flat to the retiree may well have been higher than the current intrinsic value to her, because of the social and/or physical changes that have occurred in the interim. We can imagine that the real value to her today might well be less than it was previously. Indeed, she tnight be willing to sell out to those who, because of those changes, now place a higher nominal market value on the flat than it had when she originally bought it. She could just sell for a higher than original money price, which, in the eyes of SHS, should R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . 14 SAJEMS NS Vol 4 (2001) No I attract a capital gains tax if not exempted on compassionate or administrative grounds. Currently, that is currently the case for the ZA CGT proposal if this flat is her principal residence and if her money "gain" does not exceed one million rand (This is approximately the current value of middle-class housing in Cape Town. Eventually, inflation will create many more CGT payers among those who now imagine themselves exempt). However, selling may not be a realistic consideration for her unless she can buy alternative accommodation that is at least as satisfactory as currently. Indeed, unless she can use the proceeds of her sale to buy accommodation that is at least as satisfactory as originally, she is not, in reality, better off for having experienced a measured capital gain. Putting her back in her original position with respect to the intrinsic value of her housing services may be quite expensive - especially, if it entails moving offshore in the face of capital controls on foreign exchange. If a capital gain tax is levied on the difference between the nominal purchase and sale prices, it decreases the ability of the retiree to maintain her capital assets at present real value, much less to restore that value to its original level. This may well prevent a sale that would be efficient from the standpoint of re- allocating assets from less to more efficient use. At the extreme, it may forestall the sale forever, or until the death of the retiree and her heirs. This is one example of the so-called "lock-in effect" of capital gains taxation. Example 2: The same scenario could be played out for a business, which uses an asset (say, business premises in Hillbrow), in the production of real income in the value-added sense (say, selling books on a particular religion popular in Hillbrow at that time). Over the years, changes in the intrinsic nature of that asset made it less appropriate for this specific business while making it more appropriate for other businesses. Consequently, while the market price of this asset may have risen, its intrinsic value to the business firm falls relative to its original value, which justified its initial purchase. However, an ongoing business firm would only sell this asset if it could be replaced with one of at least the same current value. OthelWise, the owner(s) would not gain from the sale relative to his (their) current position (as bad as it may otherwise be, moving may only make things worse). The fU111 would only regain its original real position if it could sell the asset in question for a high enough price to replace it with an equivalent asset, in this case, a business location elsewhere in a community that would re-establish its original profitability. Only if the current market price were high enough to finance better premises would there have been a real capital gain for the owner(s). Again, taxing the nominal capital gain will simply forestall an otherwise appropriate re- R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . SAJEMS NS Vol 4 (2001) No 1 15 allocation of capital. Again, the lock-in effect delivers an inefficient result that would not otherwise occur in the absence of CGT. 3.3 Market Versus Personal Gains: Notice that the above examples stressed a change in relative (that is, real, not necessarily nominal) market prices, which moved in the opposite direction to that of intrinsic or real value to the original owner, because of events external to the owner. That is, the owners in both cases experienced what economists call negative externalities. Generally, economists, and sometimes others, are sympathetic to those who suffer externalities - at least to the extent of agreeing that they should not suffer further as the result of government action. Indeed, if anything, they would urge the government to compensate those sufferers. This would especially be the case if the government had played any role in creating conditions promoting these negative externalities (such as relaxing zoning restrictions and crime prevention). Otherwise, economists would say that the government should tax, or the individual( s} should sue, those who are responsible. Indeed, compensation paid to those who suffer, as with an insurance or tort payout, that essentially restores them or their damaged property to its original condition, escapes income taxation, as do depreciation allowances intended to restore the original condition of the asset when renewed or replaced. The ZA-CGT proposal, in fact, will also not tax such compensation. However, economists' sympathies might be different if the intrinsic value of an asset changed due to a change in the owners' taste or production possibilities, since to some extent, the owner might well be considered responsible - even though the same principle concerning real personal gain might apply. Example 3: Consider an actual, but little known, South African case. It did not go to court and it was not publicized. The tax authorities eventually saw the "light of the train headed down the track" at them if they persisted - and the prospect was not particularly pleasant: A wealthy, but "culture-impaired", individual started to develop a taste for art. At first, her taste was not very good, in absolute or subsequently self-evaluated terms, and the art initially purchased was eventually sold in a market flooded by the desires of others similarly newly endowed with (mediocre) art appreciation and enabling means. R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . 16 SAJEMS NS Vol 4 (2001) No 1 As the tastes of our protagonist changed (improved), her previously collected works lost their intrinsic value for her. From time to time, she would sell these works whenever she was able to get prices that enabled her to purchase new works to restore her initial intrinsic value of her collection. Over the years, the gains from these sales mounted, as did the quality of the art on her walls, until the attention of the Receiver of Revenue was provoked by a disgruntled snitch. However, a very clever legal scholar was eventually able to convince the authorities that it was not in their interest to pursue a case where the individual in question had not made any gain in real value. Instead, she had merely substituted from what the public (she) now valued more (less) to what the public (she) now valued less (more). Thus, it could not actually be proven that she made any real gain that warranted attracting a tax, either in terms of her own taste or in terms of the taste of current market participants. 4 While correct in principle, this perspective would gain little widespread sympathy - especially from those "Robin Hoods" determined to redistribute from the "haves" to the "have-nots". However, all the above examples support the idea that a real capital gain is not the difference between the original price and the current sale price. Rather: Real Capital Gain is the difference between the current price of the asset sold and the current purchase price of the alternative, currendy available asset(s) that would currendy make the individually equally weD off as originally. Since the latter would be difficult for a tax assessor to determine with certainty, the "precautionary principle" (so popular these days with envirorunental groups) would require not taxing rather than risk inflicting uncompensated damage. 3.4 Nominal Capital Gains More sympathy can be generated for those who lose real value due to a change in the general price level or exchange rates (in the case of foreign investors) ... This is so well known and widely appreciated - even by tax specialists that advise the goverrunent - that it need not be laboured further here. Unfortunately, the concern and advice seldom affects the tax code on capital gains. Specifically, it virtually never leads to tax code provision of full relief. This fact might be taken as prima facie evidence of the hypocrisy of those who legislate, design, and write the tax code. They seem to have little apparent interest in establishing horizontal equity or vertical equity for real persons by being consistent on the appropriate margin. Rather, they pay lip service to those ideologies in order to justify another revenue grab. R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . SAJEMS NS Vol 4 (2001) No 1 17 Of course, that is exactly a major problem with the current ZA-CGT proposal - there is only the slightest recognition that gains may be only nominal, not real, and there is no bow to indexing of any kind. There is only the hypocritical, but untrue, rationale that at the ZA-CGT's proposed low introductory tax rate and the gradually lower inflation rates, this will not be an important matter of equity or efficiency. Not that there is anything unusual about such subterfuges. This is normal and predictable behaviour from a Public Choice perspective. However, foreign investors will likely be fmther discouraged by the taxation of domestic appreciation in rands that merely maintains an asset's foreign currency value. Such taxation can too easily be avoided by not investing in South Africa. Fortunately, at the moment, foreign investors are exempt from the ZA-CGT proposal. Nevertheless, domestic investors who intend to migrate, or to send money to their family members already established elsewhere, cannot be indifferent to the ZA-CGT introduction and evolution. 3.5 Real Capital Gains Consider some examples of how real capital gains arise in a modern economy. 3.5.1 Flow-Based Gains: I) Current-period, formal-sector income could be saved, rather than consumed, and invested in formal-sector assets at an internal rate of return equal to the going external capitalization rate for that class of assets. The expected stream of future earnings will increase and its capitalization will result in an appreciation of the assets' value by an amount equal to the saving (investment). Taxing this gain is equivalent to attempting to tax income at a higher rate if saved (invested) than if consumed. A growth-oriented government would presumably want to do exactly the opposite. la) Suppose the income came from the informal or underground sector. Then, taxing that gain, taxes only saving (and productive investment), but not consumption, from that otherwise untaxed income. Some extra revenue might be collected in this way. lb) However, it might also provide an incentive to reinvest that income in informal sector assets, that is, in assets whose capital gain and subsequent future stream of earnings might fmther and forever escape the taxman's attention. Potentially. more revenue could be lost thereby. R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . 18 SAJEMS NS Vol 4 (2001) No 1 Taxing the future stream of formal-sector earnings taxes the original saving (investment) indirectly, even with avoidance. 2) Alternatively, the investment might be made at a higher (lower) internal than external rate of return. That would lead to an even higher (lower) stream of expected future earnings and, possibly a higher (lower) capitalization multiplier, hence, yielding a gain (or even a loss) in value greater (less) than the amount of saving (investment). Taxing this extra gain (or refunding the loss) is equivalent to taxing efficiency-increasing investing (subsidizing efficiency-decreasing investing). Again, a growth-oriented government would want to do exactly the opposite! 3) Individuals may reduce their market participation, or their leisure, to devote time to augmenting their human and non-human capital by (say) studying, creating an interesting social and artistic environment (improving personal growth), creating a viable business plan, and/or remodeling or building a primary (vacation) home (improving personal well-being). Market or non-market income may be forgone or reallocated to other non-market use - thereby, for the market income portion of opportunity cost, escaping direct taxation. Thus CGT may also be avoided if this investment activity is directed primarily at augmenting human capital for which a capital value is not normally available. While taxing the resulting gains on business or select personal assets might capture the successfully invested portion of such non-market income, one might well question the incentive effects of such taxation. An extreme form of such an approach was observed in the former USSR where subjects tended to be relatively "human-capital rich" and "non-human capital poor" because the lack of effective property rights in non-personal assets constituted a 100% tax rate on capital accumulation. 3.5.2 A Misleading Fiction: Further, a common, but misleading, fiction (or myth) has been maintained in cases 1-3 directly above. The fiction is that capital gain can arise from current income. The examples show how saving from current market or non-market income might compare with a change in an asset value. But whether there is any capital gain depends, not on current income, but on a current investment decision, which given the incentives, is made more or less successfully or unsuccessfully. R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . SAJEMS NS Vo14 (2001) No 1 19 That is, it is the decision-making, and, specifically, the desire and ability to do it well, that is the key to capital gain or/oss, not current income. 3.6 Taxing Tbe Future From a forward-looking perspective, so-called capital gains are always capitalized expected future income gains, not present income. Taxes on capital gains are taxes on future income, not present income - not, at least, the sort of present income that most public finance specialists say they are worried about avoiding income taxes. Rather, it taxes productive (as opposed to unproductive) investment decisions, which for the most part, are based on information whose production generally does lead to taxable income in the current or previous periods. Economists have never argued that decision-making per se (such as deciding to consume or invest in X rather than Y) should be taxed. From economists' own cultural perspective, good decision-making creates more value-added than bad decision-making. Those who make more good decisions than bad may well earn a market or implicit fee or income for such services. But typically, when such commissions are explicit they are taxed in the present, and they generally do not represent more than a small fraction of the gains expected to be created thereby. The latter is value-added expected, and quill! properly taxed, in the future. Thus, if the goal is to tax current income, which would otherwise escape taxation in the current (or previous) tax year(s), capital gains should not be taxed. In summary, the reasons are as follows: 1) There is no necessary reilltionship between capital gains and current income otherwise escaping current taxation. 2) Capital gains taxation actually taxes good decision-making with respect to creation of future taxable income - thereby potentially decreasing the future tax base, as weB as future income and the future capital stock. 3) Increments to future income, due to increments in real non-human and human capital, are taxed in the future. From a forward-looking perspective, a CGT is simply a special type of "Transactions Tax". That is, it is a tax on transactions in marketable assets which have been held for longer periods than the limit (generally one to two years)5 set in the tax code as determining such transactions as part or a normal, for-profit business operation. Indeed, although it has often been called a voluntary tax (because it can be avoided by avoiding the transaction), from a forward-looking perspective, this is a bit of a misnomer. From the case studies presented above, and a bit of knowledge about the determinants of business and R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . 20 SAJEMS NS Vol 4 (2001) No 1 personal choice, we can see that the tax leads to an involuntary postponement of a transaction, which might otherwise have been optimal to pursue sooner. 3.7 Systemic Problems Usually, public fmance specialists advocate taxation of corporate income and capital gains as a way to insure and protect the revenue productivity of the personal income tax. They fear that those who have an unequal ability to shift their de facto current income to the de jure havens of corporations or capital assets (if untaxed), will have an unequal advantage in avoiding personal income taxation. Such shifts, on a large scale, might imperil the personal income tax base and, hence. revenue collections. Such fears are ultimately due not to potential taxpayer choices but rather to legislative and administrative choices made over base definition and differential tax rates on different bases. 3.7.1 Tax Base: First. consider the primary problem of base specification. where we meet again the ugly. out of date, and confusing role played by the SHS concept of income. A Canadian example may again be instructive. Under the Carter Commission conception of SHS, the corporation was not an appropriate object of income taxation theoretically - only practically. For Carter, only real persons enjoyed SHS income. Income earned by corporations. according to Carter, should be attributed to owning individuals. where taxes on it are properly paid at the marginal tax rate applicable to each individual. respectively. Thereby, horizontal and vertical equity could be easily achieved at one stroke for each individual income recipient - presuming. very cavalierly, that there was no tax shifting. At most, corporations should just serve to as withholding entities for capital income, just as they were already doing for labour income. The beauty of this conception of an income tax system was truly astounding - at least to the Carter Commission and its fans (at the time, including me). The Canadian politicians thought otherwise and delivered their own extensively revised version of the Carter Report nearly four years later in Finance Minister Benson's 1971 White Paper. Even that was subsequently further revised before it emerged as legislated tax reform in 1972, which retained separate, un-integrated taxation of corporations and introduced a capital gains tax to wary Canadian taxpayers and ecstatic Canadian tax accountants (who thereby, saw their consulting income prospects skyrocket!). The Carter recommendations simply proved to be too impractical both for Revenue Canada and for the Liberal Government. The ideological. if R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . SAJEMS NS Vol 4 (2001) No 1 21 not practical, thrust of SHS was retained, however, because some rationale was needed to justify refonns and especially the introduction of a capital gains tax. The end result of attempting to improve consistency along the SHS-Carter model was to make the tax system differently inconsistent. By attempting to eliminate avoidance more opportunities for avoidance were created. While an integrated tax base would have eliminated a certain type of avoidance, it was not politically or practically feasible. 3.7.2 Tax Rate: Second, given the non-integration of corporate and personal income, there is the progressive rate structure problem. A capital gains tax on corporate equity (but not on non-corporate assets) might be justified as a way to reduce avoidance of personal income taxation by the route of corporate retained earnings. Whether this is a substantial problem that needs to be addressed in this way depends crucially on a comparison of the top personal and corporate marginal tax rates. If they are the same, or close, there is no or little incentive to use the corporate form to avoid tax, 6 and hence, little justification for a capital gains tax specifically on corporate owners or on any asset or anyone else, for that matter. Of course, this problem would be easily resolved by dropping the ambition and pretence of achieving vertical equity. Then, the progressive rate structures could be flattened or even eliminated. That would reduce or eliminate the main reason that taxpayers might want to substitute between tax bases.7 Indeed, a flat tax rate structure would reduce or eliminate many of the problems inherent with income taxation, along with their partial, prospective solutions such as CGT, income averaging, dividend credits, tax credits versus deductions, etc. (Hall & Rabushka, 1995). It would also eliminate the need for bloated tax collection bureaucracies and the diversion of scarce accountants, and less scarce lawyers, into the tax-avoidance industry. Strangely, orthodox public finance specialists seldom consider this route to "consistency". 4 CONCLUSION In a sense, South Africa is starting with an almost clean slate with respect to certain kinds of income taxation. It should strive to keep that slate as clean and simple as possible. In the name of consistency, the introduction of capital gains taxation in any form, will multiply, not reduce, the system's inconsistencies. In the name of equity, it will multiply the inequities. It will not raise substantial net revenues. But it will be a source of ongoing lobbying activity, and that will undermine social cohesion. R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . 22 SAJEMS NS Vol 4 (2001) No 1 In the short l1lIl, ZA-CGT may yield political and financial benefits for the governing party. In the long run, if that party's constituents' market prospects improve as a result of other government policies and/or market developments, its own prospects are likely to suffer. Every time a constituent's income earning or wealth gaining prospects are blocked or reduced by the presence of a CGT, votes and financial support wiH be threatened. The "fmancial cows" will increasingly avoid that party's "milk house'" Other parties will offer alternatives in order to farm this ever more fertile avoidance field. The prospects for the economy will be diminished by the incentives to divert resources from market to political opportunities. ENDNOTES Thanks are due to D. Allen, J. Chant, H. Grubel, D. Hammes, R Jones, R Krelove, and J. Stuart. 2 Notwithstanding the following: a) Most of the goods provided by modem governments are not pure public goods (See Blomquist & Christiansen, 1999, for a rationale). b) Modem theoretical analysis, experimental results, and real world observation suggest that the free riding hypothesis does not strongly apply in practice or can be managed privately with technological or contractual fixes (For example, Rondeau et al., 1999). c) Government itself is a public good, which under the strong version of the free-rider hypothesis would not be spontaneously provided (That is, the free-rider hypothesis undermines evolutionary explanations for the existence of government). d) Modem Public Choice analysis suggests that government provision promotes rather than reduces free riding (De Jasay, 1989). 3 These are not uncontroversial results. See Lee (1995) for some counter· arguments with respect to the US. It rnight be noted in connection with his critique and in anticipation of the arguments in the next part, that the theoretical results apply to changes that are expected to be permanent. However, few rational political and financial observers of the US would regard any reform as more permanent than the composition of the· legislature that passed it or the special-interest groups that supported it. 4 This example, with a slight change in identity to protect or impugn the innocent, was told to me by C. Prlsman of Cape Town, who further states: "Of course, the revenue department was seeking to tax income and to treat the whole venture as a profit making exercise - but how could it be? Taste changed and taste is not a venture. When once the intention was to acquire permanently to enjoy - value was irrelevant. As enjoyment ceased, what to do then? Well to buy what you now enjoy is the reason for the sale, not the issue of obtaining income." R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . SAJEMS NS Vol 4 (2001) No I 23 ceased, what to do then? Well to buy what you now enjoy is the reason for the sale, not the issue of obtaining income." An extension of Prism an's idea was given me by David Hammes: Suppose you have a "patron" who likes to buy work by new, up-and-coming artists .. .if slhe cannot sell their "old" paintings and buy new ones repeatedly - (the effect of imposing a capital gains tax is to keep assets whose returns would be taxed "frozen") then the market for new artists shrinks substantially. Who is likely to be poorer? Older, established, artists or new, up-and-coming artists? So, the CGT actually is borne by the poorer part of the population. 5 South African tax law (section 9B) provides that under certain conditions the proceeds of the sale of listed shares held for a period of at least five years may be regarded as of a capital nature. See Katz (1996). 6 When marginal rates were in the 90% range while corporate rates were in the 50% range, as was the case in the USA for quite a while, the incentive was very large and widely experienced. With top marginal and corporate rates within around 10 percentage points of each other, as is currently the case for South Africa, the incentives, and those motivated by them, are probably not great enough to justify a special legislated tax, as opposed to tax code procedural changes. 7 As far as I know, public fmance has yet to justify scientifically any specific progressive rate structure guaranteed to deliver vertical equity in practice. (Most recently, Plug et al. (1999) made a heroic, efficiency- preserving attempt based on the assumption that everyone has the same utility function of a particular form.) Actual progressive rate structures are politically determined by bargaining with affected lobby-groups under the justifying general interest ideology of vertical equity. The latter, to a professional con man, would simply be known as "the hook", which is the premise on which any successful con game must be based. REFERENCES ATKINSON, A., CHARI, V.V. & KEHOE, PJ. (1999) "Taxing Capital Income: A Bad Idea", Minneapolis FRB: Quarterly Review, 23(3): 3-17. 2 BLACK, P. (2000) "Capital Gains Tax: Critical Notes", The South African Journal of Economics, 68 (4): 805-9. 3 BLOMQUIST, S. & CHRISTIANSEN, V. (1999) "The Political Economy of Publicly Provided Private Goods", Journal Of Public Economics, 73 (1): 31-54 4 CHANT, J. (1999) "Two-Handed Economists: Do We Practice What We Teach?", The Fraser Institute Symposium on Capital Gains Taxation: Vancouver. R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . 24 SAJEMS NS Vol 4 (2001) No 1 6 DE JASAY, A. (1989) "The Return of the Free Rider", In his Social Contract and Free Ride, Claredon Press: Oxford, 219-44. 7 GRAD STEIN, M. (1999) "Optimal Taxation and the Fiscal Constitution", Journal of Public Economics, 72 (3): 471-85 8 GRAHAM, J. R. (1999) "Do Personal Taxes Affect Corporate Financing Decisions?", Journal of Public Economics, 73 (2): 147-85. 9 GROTE, M. and FLETCHER. K. (2000) "Capital Gains Tax in South Africa", The South African Journal of Economics, 68 (4): 786-96. 10 GRUBEL, H. G. (2000) Unlocking Canadian Capital: The Case for Capital Gains Tax Reform, The Fraser Institute: Vancouver. 11 HAIG, R. M. (1921) "The Concept of Income - Economic and Legal Aspects", In R M Haig, (ed.), The Federal Income Tax, Columbia University Press: New York. 12 HALL, R and RABUSHKA, A. (1995), The Flat Tax, Hoover Institution Press: Stanford. 13 HINDRIKS, J. (1999) "The Consequences of Labour Mobility for Redistribution: Tax vs. Transfer Competition", Journal Of Public Economics, 74 (2): 215-34. 14 HINES, J. R. Jr. (2000) "What Is Benefit Taxation?", Journal of Public Economics, 75 (3): 483-92. 15 JONES, R. W. (2000) Globalisation And The Theory of Input Trade, MIT Press, Cambridge. 16 KATZ, M. (1996) Katz Commission Report, bttp"//www \lOY zalrcports/1996/katz bt. 17 KNIGHT, B. G. (2000) "Super-Majority Voting Requirements for Tax Increases: Evidence from the States", Journal of Public Economics, 76 (1): 41-67. 18 KNELLER. R., BLEANEY, M. F., & GEMMELL, N. (1999) "Fiscal Policy and Growth: Evidence from OECD Countries", Journal Of Public Economics, 74 (2): 171-90. 19 LANG, M H & SHACKELFORD, D. A. (2000) "Capitalization of Capital Gains Taxes: Evidence from Stock Price Reactions to the 1997 Rate Reduction", Journal of Public Economics 76 (1): 69-85. 20 LEE, J. (1995) "Capital Gains Myths", Tax Notes, (May 8). 21 Minister of Finance, web site: www.fmance.gov.za. 22 PLUG, E. 1. S.; VAN PRAAG, B. M.S. & HARTOG, J. (1999) "If We Knew Ability, How Would We Tax Individuals?" Journal Of Public Economics, 72 (2): 183-211. 23 RONDEAU, D.; SCHULZE, W. D. & POE, G. L. (1999) "Voluntary Revelation of the Demand for Public Goods Using a Provision Point Mechanism", Journal of Public Economics, 72 (3): 455-70. 24 SIMONS, H. C. (1938) Personal Income Taxation, University of Chicago Press: Chicago. R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . SAJEMS NS Vol 4 (2001) No 1 25 25 SMETTERS. K. (1999) "Ricardian Equivalence: Long-run Leviathan". Journal Of Public Economics. 73 (3): 395-421. 26 STEENEKAMP. T J. (2000) "Good Tax Practice and Taxing Capital Gains''. The South African Journal of Economics, 68 (4): 798-804. 27 TULLOCK G. (1981) "The Rhetoric and Reality of Redistribution". Southern Economic Journal, 895-907. R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) .