1 4089 A SIMPLER CONSUMPTION-BASED ALTERNATIVE TO THE INCOME TAX FOR SOCIALIST ECONOMIES IN TRANSITION Charles E. McLure, Jr. The tax systems of socialist economies in transition will distort resource allo- cation, create inequities, and cause administrative headaches if not reformed. These countries have weak tax administrations, lack experience with mass taxes based on voluntary compliance, and need to encourage domestic saving and foreign investment. This article suggests an alternative to the conventional income tax that is more suited to these conditions. Attempting to tax real economic income raises complicated timing issues (when to recognize income and allow deductions) and may require complex adjustments for inflation. The simplified alternative tax (SAT) avoids these com- plications and provides a general incentive for saving and investment less sub- ject to abuse or distortions than tax holidays and other tax gimmicks in vogue in countries emerging from socialism. The key elements of the SAT are separate taxes on income from labor and capital, immediate deduction for all business expenditures, no deduction for interest, and no taxation of interest or dividends. (Interest could be treated as under an income tax, at some cost.) Although the marginal effective tax rate is zero, the government shares in extraordinary returns to investment. The ar- ticle discusses potential problems (including distributional implications, tax losses, and foreign tax credits) as well as advantages of the SAT. C ountries emerging from socialism need new tax systems that will not distort resource allocation, create inequities, or cause administrative I ~~ overload. Most of these countries are adopting the tax models of The World Bank Research Observer, vol. 7, no. 2 (July 1992), pp. 221-237 3 1992 The International Bank for Reconstruction and Development/THE WORLD BANK 221 advanced Western economies-a value added tax (VAT), income taxes on en- terprises and individuals, and excises on a limited number of products (see Gray 1991 on Czechoslovakia, Hungary, Poland, and the former Soviet Union; McLure 1991 for Bulgaria). Some are rushing to adopt tax incentives, despite a worldwide movement away from that strategy. This article suggests an alternative to the conventional income tax-one that is more suited to countries with weak tax administration and no experience with mass taxes based on voluntary compliance and self-assessment and also meets the need to encourage domestic saving and foreign investment. The sim- plified alternative tax (SAT) would substitute for corporate and individual in- come taxes and might replace the VAT and payroll taxes as well. This article contrasts the complexity of the income tax with the simplicity of the SAT and describes both the economic advantages and the potential problems of the SAT. One legacy of socialism is lack of experience with tax administration and compliance. Under central planning, profits from state enterprises ("funds" with no other use) were transferred to the budget through bookkeeping entries. De- fining or measuring income correctly was not important. The base of profits taxation, "balance sheet profits," bears little resemblance to economic income. (It allows no deductions for depreciation or for some interest expenses.) The turnover taxes of the reforming economies, which are simply the difference be- tween two controlled prices, require no real administrative or compliance effort. The situation will be quite different after the transition to a market economy. In The Road to a Free Economy, Kornai (1990, pp. 118-19) notes another legacy of socialism, fear of the state and disrespect for its laws: People in general consider it a laudable act, rather than something to be ashamed of, if someone defrauds the state, appropriates its wealth, or shuns his own obligations. Those who refrain from this kind of behavior are seen as dupes.... Consequently, when we con- template budget revenues we should be prepared to face the fact that many citizens will try hard to dodge taxes.... In the Hungarian case, there is an additional factor: a considerable part of the private sector still belongs to the shadow economy, and it will voluntarily emerge from the shadow and into the light only after some time. Kornai rightly fears a system in which tax collectors spy on taxpayers. In short, every effort should be made both to upgrade tax administration and compliance and to design tax policies with the constraint posed by inade- quate administration and compliance firmly in mind, using withholding when- ever feasible. The SAT, unlike the conventional income tax, facilitates meeting the second goal. While the SAT promises similar advantages for developing countries (see McLure and others 1990), it deserves special attention in coun- tries emerging from socialism. Their tax administration and compliance are es- pecially weak, and they do not have the option, as developing countries do, of continuing to use their present tax systems; they must reform them. 222 The World Bank Research Observer, vol. 7, no. 2 (July 1992) Complexity of the Income Tax For an income tax to be fair and neutral, its base must closely resemble real economic income. Meeting this objective is not easy. Timing issues are inevita- bly complicated, and complex adjustments for inflation may be necessary in measuring income. Timing Issues Taxable income understates economic income in present value terms if de- ductions are allowed "too soon" or income is recognized "too late." Deprecia- tion is the best-known timing issue. Because it is hard to know precisely how rapidly assets lose value, depreciation for tax purposes is commonly based on arbitrary schedules. If these deviate too far from reality, equity, the perception of equity, and economic neutrality suffer. Administrative simplicity may also be a casualty of seemingly simple but overly generous depreciation allowances, as new limits and more complex rules are imposed to prevent perceived abuses. Multiperiod production occurs when expenditures are incurred in more than one year for the production of income that may also be realized in more than one year. Classic examples are timber, other trees with long gestation periods and productive lives (such as coffee, rubber, and fruit trees), mine development costs, and large construction projects. The conceptually correct approach under an income tax is to recognize the annual increase in the value (of the mine, dam, or tree) as production occurs and to allow current deduction of expendi- tures. This method is often impractical, however, because of valuation and li- quidity problems and political opposition. Imagine taxing the increase in the value of coffee trees or grape vines during their long gestation periods. A conceptually inferior but more practical alternative is to postpone recog- nition of income until income-earning activity begins, to capitalize expendi- tures, and to allow depreciation over the productive life of the asset. This alternative is not without its own difficulties, from the amount of accounting required to the question of which costs should be capitalized and which de- ducted currently. These problems also arise in accounting for inventories. For long-term contracts, interim tax liabilities may be based on estimates of total income and the percentage of the contract completed each year, with retroac- tive adjustments at the end of the contract. The conceptually correct treatment of capital gains-current accrual taxa- tion-is administratively and politically difficult (valuation and liquidity prob- lems). To avoid these difficulties, capital gains are commonly taxed when assets are sold. By postponing liability, this treatment undertaxes gains, undermines horizontal and vertical equity, and reduces the mobility of capital. Under a graduated tax rate schedule, the bunching of gains in the year of disposition can cause overtaxation. Averaging gains is complicated, and preferential treat- Charles E. McLure, Jr. 223 ment of capital gains is not a satisfactory solution to this problem-or to in- adequate capital formation or the need for inflation adjustment. Original issue discount occurs when a bond is issued at less than its redemp- tion value; interest is implicit in the increase in the bond's value as maturity approaches. Assuming that the increase occurs at maturity undertaxes this form of income. Because of compounding, assuming that interest is earned pro- portionately over the period from issue to maturity errs in the other direction. Interest is earned as indicated by a table of compound interest, and to assume otherwise produces an inaccurate measure of income. This is a sample of timing issues. A complete list would include such knotty cases as installment sales, decommissioning of nuclear power plants, and rec- lamation of strip mines. Common factors among all timing issues are uncer- tainty about when income is earned or when a change in value occurs; complexity in accounting, compliance, and administration; and potential dis- tortions and inequities. Timing issues are inevitable in an income tax. Even if the esoteric problems can be avoided, many of the mundane ones will remain, and others will surface as taxpayers and economies become more sophisticated. If they are not handled satisfactorily, equity and neutrality suffer. Problems arise if interest is not treated symmetrically, so that it is deductible by the debtor and taxable to the creditor in the same year. For example, if tax- payers can choose their fiscal years for tax purposes, they can manipulate loans between related parties (say, a partnership and its partners) to deduct interest before it is taxed. The result is inequity and revenue loss. If deductions are sufficiently accelerated, the recognition of income is suffi- ciently postponed, capital gains are taxed preferentially, or important transac- tions are not treated symmetrically, taxpayers may report losses for tax purposes, even if they are making money. If business losses are used to offset or shelter income from other sources, the perception of inequity may create political problems. "Backstop" measures used to assure that wealthy individuals and large and profitable companies pay some tax can add greatly to complexity. Inflation Adjustment Most tax systems assume stable prices. They use historical costs to calculate capital gains, depreciation, and the cost of goods sold, and they do not distin- guish between nominal and real interest payments. Even at modest levels of in- flation, it is necessary to choose between the complexity of inflation adjustment and the inequities and distortions of an unindexed system. Several examples show the need for inflation adjustment. Consider capital gains. If the price level has doubled since an asset that sells for 160 was bought for 100, there is no real gain; there is a real loss of 40 (160 - [100 x 2]). Yet taxation based on his- torical value reports a gain of 60 (160 - 100). Inflation adjustment of the basis of assets produces the correct result. 224 The World Bank Research Observer, vol. 7, no. 2 (July 1992) Depreciation allowances create similar problems. If the price level has risen significantly since a depreciable asset was acquired, it is appropriate to base depreciation deductions on the inflation-adjusted value. Use of historical value understates the loss in real value and overstates taxable income. Depreciation is sometimes accelerated to offset inflation. But a given pattern of acceleration is appropriate for one inflation rate at most. Inventories can also be indexed for inflation. Some countries allow last-in, first-out (LIFO) inventory accounting, which bases the cost of goods sold on the cost of the most recent purchases, as a substitute for inflation adjustment. This is not conceptually correct, for it confounds changes in relative prices of inven- tories and changes in price levels (the proper object of indexation; for further discussion, see McLure and others 1990, ch. 7). Consider finally interest expense and interest income. Suppose the real in- terest rate is 5 percent, inflation is 20 percent, and the nominal interest rate is 25 percent. (Fisher's Law says that in the absence of inflation, the nominal in- terest rate is approximately equal to the sum of the inflation rate and the real interest rate. Modification for taxation would not change the basic argument.) A deduction for (and tax on) only the real interest payment of 5 percent is ap- propriate; the rest of the nominal interest rate represents compensation for loss in the value of principal. This example shows that failure to adjust for inflation as low as 20 percent can have a major impact on tax equity and neutrality. The examples suggest one form of inflation adjustment: independent adjust- ments of the four affected income flows (capital gains, depreciation, invento- ries, and interest). An alternative integrated approach is more accurate; it adjusts nonmonetary elements of the balance sheet and then reflects the adjust- ments in the income statement. Adjustment of interest is an implicit result of the process. (McLure and others 1990, ch. 7, present a full discussion.) Both types of inflation adjustment are complicated. Either would be a major burden for any country-doubly so for a country with rudimentary accounting, weak tax administration, unreliable price indexes, and no experience with taxpayer compliance. Socialist economies in transition should not undertake inflation adjustment gladly or lightly. Whether these countries need inflation adjustment is unclear. Historical ex- perience provides no guidance; it reveals only the artificial price stability of sup- pressed inflation. Inflation may be fairly high in the years following tax reform. If so, reliance on an unindexed tax system would be risky. Incentives for Saving and Investment To encourage domestic saving and investment by residents and foreigners, many countries emerging from socialism are adopting a "Swiss cheese" strategy of targeted investment incentives, tax holidays, and other fiscal inducements, rather than adopting a comprehensive definition of taxable income and relying Charles E. McLure, Jr. 225 on low tax rates and market forces, as many western industrial countries have recently done (see Boskin and McLure 1990). This strategy is undesirable. Its equity is questionable, and it creates a strong impression of inequity. Higher tax rates are required than under a comprehen- sive income tax. Opportunities for abuse strain the ability of weak tax admin- istrations. These include manipulation of transfer prices on sales between affiliated firms, and the "yo-yo" tactic of sending money out of the country so it can return as tax-preferred foreign investment. The strategy implies an unjustified faith in bureaucrats and politicians to pick winners and losers. The results of central planning do not justify this faith. If the wrong sectors are chosen-or choose themselves, via abuses-scarce re- sources are wasted. Generalized and neutral incentives are more appropriate. Dividend Relief Under classical systems, the equity income that companies distribute to shareholders is taxed twice. The company pays tax on the income and then individuals pay tax on the dividends. Since interest is deductible and dividends are not, such systems encourage debt finance and recharacterization of divi- dends as interest. There may be an incentive for businesses to avoid the corpo- rate form (see McLure 1979). Most Western European countries allow relief from double taxation of dividends, commonly by granting shareholders a tax credit for company tax that is attributable to distributed income. A deduction for dividends paid appears simpler, since it lodges compliance problems at the enterprise level. But if there is a withholding tax on dividends, the two are ad- ministratively equivalent. Exemption of dividends from shareholder taxation, while not conceptually sound, may be appropriate for socialist economies in transition (see McLure forthcoming). Deviations of taxable from economic income complicate dividend relief. Most countries limit relief to company taxes paid. Such provisions are compli- cated and can distort decisionmaking (McLure 1979, ch. 4). Intercorporate div- idends are another source of complexity that could be especially important in transitional economies. Holding companies may be established during privati- zation, intercompany holdings will probably be common, and there are likely to be many joint ventures. In sum, avoiding double taxation of dividends is complicated. It is not something that countries emerging from socialism will soon be prepared to handle. Global Taxation and Withholding One of the widely accepted tenets of income tax policy is that global taxa- tion (taxing aggregate income under a single rate schedule) is conceptually pref- erable to schedular taxation (taxing specific types of income under separate 226 The World Bank Research Ohserver, vol. 7, no. 2 (July 1992) schedules) on grounds of both equity and neutrality. But schedular taxation can be attractive for reforming socialist economies. It can prevent the offsetting of artificial tax losses against other income and can simplify withholding and re- duce the number of taxpayers who file returns. Under a global income tax with graduated rates, a taxpayer's tax liability depends on aggregate income. Withholding or informational returns may be necessary to prevent evasion of tax on interest and dividends. If schedular taxes are applied to interest and dividends (or if they are exempt), withholding is simplified. Allowing deductions for interest payments of individuals, including mortgage interest, also complicates matters. The Simplified Alternative Tax The SAT, though resembling a standard income tax, differs in ways that have both administrative and economic importance.2 How It Works The SAT consists of two schedular levies: one on labor income and one on business and capital income. The tax on the labor income of individuals, in- cluding pensions, can be collected almost entirely through source withholding if it is simple enough. The individual SAT can allow personal exemptions and graduated rates, but it need not. It can accommodate itemized deductions and joint filing, but to facilitate withholding, these complications should be avoided initially. (Most planned economies have no tradition of joint filing and itemized deductions.) Gifts and inheritances could be subject to the indi- vidual tax. The novelty of the SAT is its treatment of income from business and capital. All such income (except pensions) is excluded from the base of the individual tax. Business income is subject to a separate, flat-rate tax, whether earned by a corporation, state enterprise, joint venture, collective, partnership, or propri- etorship; the business tax rate is usually assumed to equal the top individual rate. Business losses cannot offset individual income, but owners of small busi- nesses (including cooperatives and corporations) can pay salaries to themselves to achieve "do-it-yourself integration." Expensing (deduction in the year of expenditure) is allowed for all investments, including depreciable assets and inventories. Interest and dividends are not deductible expenses, and they are tax-exempt, whether received by individuals or businesses. Borrowing and lending have no tax implications.3 Capital gains on financial assets are exempt. Gains on nonfinancial assets are exempt from the individual tax but subject to the business tax. Individuals buying or selling nonfinancial assets would file business returns. In simple cases these could be schedules filed with the individual tax. Charles E. McLure, Jr. 227 Most authorities would treat owner-occupied housing like financial invest- ment. Purchase or sale of a home would have no tax implications; home mort- gage interest would not be deductible. Its Simplicity The SAT avoids problems of timing and inflation adjustment. Interest pay- ments have no tax consequences, and business tax is based on cash flow rather than accrual. Expensing replaces depreciation and inventory accounting; there is no need (for tax purposes) for depreciation or inventory accounting or for inflation adjustment of depreciation allowances or inventories.4 (These may be addressed in financial accounting, discussed below.) With expensing, the basis of capital assets is zero. All proceeds from the sale of assets are taxable. There is no need to account for the (depreciated) basis of assets or to index it for inflation. This feature facilitates cross-checking income tax returns of buyers and sellers. Deductions for the purchase of assets and in- ventories are contemporaneous with the reporting of receipts, rather than spread over time, as in the case of depreciation allowances and deductions for the cost of goods sold. Interest is tax-exempt and nondeductible, so timing problems such as origi- nal issue discount and inflation adjustment of interest payments do not arise. Neither postponing the recognition of income nor accelerating deductions can undermine neutrality or fairness. Because interest and dividends are treated identically, the SAT avoids differential treatment of debt and equity, without complicated provisions for integration. There is no need for withholding or informational returns on interest and dividends. Liability for the individual tax depends only on labor income and family circumstances (working spouse, personal exemptions, itemized deduc- tions), not on nonlabor income. Thus withholding can more accurately reflect final tax liability under an SAT than under a global income tax. This minimizes the number of individual returns that must be filed. Because the SAT is schedular, tax shelters based on pass-through of losses cannot exist; sheltering can occur only within a single business filing one re- turn. This benefit is bought at the expense of implicitly taxing all business in- come (except that paid to owners as wages) at the business tax rate rather than the marginal rates of the owners. It is important not to claim too much for the SAT. There is no suggestion that the tax return could fit on a postcard, as in the flat tax proposal of Hall and Rabushka (1983, 1985), but most taxpayers would not need to file returns. Taxation of individual income resembles the taxation of labor income under a standard income tax; it involves no major simplification, relative to a sensible income tax. The big differences are the schedular nature of the system, expens- ing, and the exemption and nondeductibility of interest and dividends. 228 The World Bank Research Observer, vol. 7, no. 2 (July 1992) The SAT facilitates compliance for (and administration of the tax on) honest taxpayers. It does little directly to curb certain abuses or to facilitate taxation of hard-to-tax groups. But it helps indirectly by freeing administrative resources for enforcement of the tax laws. Economic Effects of the Simplified Alternative Tax Several of the economic effects of the SAT have particular relevance to so- cialist economies in transition.5 Exemption of the Return to Invested Capital Under certain conditions, the SAT is equivalent to a tax that exempts the re- turn to investment. This is clear for interest and dividends, which are explicitly exempt. A simple example shows that other capital income is also in effect ex- empt. Consider an equity-financed investment of $1,000 that yields $200, or 20 percent, one year later. If the company tax rate is 30 percent, the company in- yests only $700 of its own money, since the deduction of $1,000 results in a tax saving of $300 (assuming the company can fully utilize the deduction, which is addressed below). Tax of $360 must be paid on "taxable income" (which ac- tually includes the return of principal) of $1,200 in the second year, leaving net after-tax income of $140 ($1,200 less $700 less $360), or 20 percent of the initial investment of $700 of company funds. Since the net rate of return on company funds is the same as in the absence of tax, the marginal effective tax rate (METR)-the percentage by which taxation reduces the before-tax rate of re- turn-is zero (see King and Fullerton 1984; McLure and others 1990). The effective exemption of the return to investment under the SAT is a pow- erful incentive for saving and investment-one that avoids the problems of tar- geted incentives: the difficulty of picking winners, the risk of economic distortions, the complexity, the opportunities for abuse, and the real and perceived inequities. Taxation of Extraordinary Returns Although the return to capital is exempt, the government shares in extraor- dinary returns to investments. In the example presented above, the state earns 20 percent on its investment of $300 in foregone tax revenues; this is true even if the competitive return is only 10 percent. The government is a silent partner in all investments. This feature could be important in socialist economies in transition. Consid- er the potential for extraordinary returns from trademarks such as Coca Cola and McDonald's, from enterprises that are privatized on too-generous terms, Charles E. McLure, Jr. 229 from inefficient but viable state enterprises that become efficient, and from the exploitation of natural resources. Parity of Taxable and Tax-Exempt Sectors Interest and dividends paid to the state are inherently exempt, and those paid to nonprofit organizations are commonly exempt. Besides lacking neutral- ity, this selective exemption invites abuse. Parity requires that interest and div- idends paid to private investors also be exempt. This implies nondeductibility of interest and dividends at the enterprise level. (Otherwise, the METR will be highly negative, considering taxes levied at both enterprise and shareholder/ bondholder levels.) This describes the SAT. Some countries exempt interest on government debt. This is an inequitable and inefficient subsidy. By exempting all interest income, the SAT eliminates the advantage of public borrowing. If business income of nonprofit organizations is exempt, discrimination is less under the SAT (which effectively exempts the return to for-profit invest- ment) than under the income tax. Potential Problem Areas When recommended in market economies, SAT proposals encounter predict- able and potentially valid objections. These objections generally appear to be less telling in the case of socialist economies in transition. Distributional Equity The effective exclusion of the return to capital under the SAT can be ap- praised in several ways. That the SAT is less progressive than an income tax levied at the same rates may be less important in economies moving away from socialism than in market economies, despite the greater weight given distribu- tional considerations under socialism.6 The concentration of wealth will de- pend on how privatization is achieved, whether through vouchers distributed to all citizens, enterprise shares distributed only to workers and managers, asset sales, or "spontaneous privatization" ("taking the enterprise private," often by means of questionable legality). A society raised to believe that profits are bad may have difficulty accepting the effective exemption of income from capital. Many enterprises with large economic incomes would pay little or no tax.7 Because interest income is ex- empt and interest expense is not deductible, the profit margin of financial in- termediaries would be exempt. A special tax might be imposed on this sector (see Kay 1988). 230 The World Bank Research Observer, vol. 7, no. 2 (July 1992) Losses Because of expensing, tax losses are more common under the SAT than under an income tax. New companies are especially likely to incur tax losses. Simple carrying forward of losses (without interest) is not satisfactory. If deductions for investments cannot be used immediately, the SAT's incentives for investment and risk taking would not be fully realized, and some simplicity would be lost. Some firms, especially rapidly growing ones, may never have the income to ab- sorb losses. (If the rate of growth of investment is as great as the rate of return, an equity-financed firm will never incur tax liability.) Carrying tax losses for- ward with interest is superior to carrying them forward without interest, but it is more complex and does not deal with perpetual tax losses. The economic benefits of the SAT could be retained if refunds (equal to the product of the tax rate and the losses) were paid currently. Refunds are com- monly opposed, however, because of the risk of fraud and the administrative burden of preventing abuse. The Cost of Debt Capital Nondeductibility of interest would raise the cost of foreign-source debt cap- ital and might be politically infeasible. Allowing deductions for interest (and presumably for dividends, to maintain financial neutrality) would, in turn, call for taxation of interest of residents (collected largely through withholding) and perhaps withholding on interest paid to foreigners. The tax would be creditable in capital-exporting countries (see discussion of international issues below; see also Sinn 1987). Such a compromise has many disadvantages. It would destroy parity in the tax treatment of state enterprises, private firms, and nonprofit organizations, and it would make tax losses more prevalent. It would also reintroduce the complexities related to timing (when to recognize interest income and the de- duction of interest expense), and it would render the tax vulnerable to inflation, unless interest payments were indexed, creating further complexity. Accounting Countries in transition from socialism have no experience with generally ac- cepted accounting practices (GAAP). The absence of GAAP hampers development and implementation of a modern income tax, as well as development of busi- ness more generally. Small businesses lack the ability to implement sophisticat- ed accounting systems. The SAT provides cash-flow treatment of receipts and expenditures and ignores interest income and expense. This simplification ad- vantage is partially offset by the need to keep dual records, one based on GAAP and one on SAT conventions. By comparison, the income tax model would en- courage more rapid development of GAAP.8 Charles E. McLure, Jr. 231 This problem of nonconformity of SAT and GAAP is easily overstated. SAT returns would be based on the same information as income tax returns, but without the complexity of timing issues and inflation adjustment. A far greater problem in many countries is the divergence between GAAP and the calculation of income for tax purposes, often reflecting deliberate policy. On balance, SAT's nonconformity with GAAP is not a matter for major concern. Manipulation of Transactions Some economically similar transactions are treated differently under the SAT (see Sunley 1988; McLure and others 1990; McLure and Zodrow 1990; Zodrow and McLure forthcoming). For example, lease payments are deductible and taxable, whereas interest is nondeductible and nontaxable. This opens the door to manipulation of transactions between two persons, one subject to the busi- ness tax and the other not. (There is generally no advantage to such manipu- lation if both parties are either taxable or nontaxable.) A taxpayer might simultaneously lease a building to a tax-exempt organization at below-market rent and borrow from the organization at a below-market rate of interest. Alternatively, the taxpayer might make an installment sale to such an organization-or to an individual-at an understated price and an overstated rate of interest. Solutions might involve floors and ceilings on interest rates or the taxation of certain business activities of tax-exempt organizations. Though the problem deserves further scrutiny, it does not appear to be important. A related and po- tentially more serious problem is overstatement of the purchase price of im- ported assets, an abuse that also is possible, though in attenuated form, under an income tax. Since an SAT allows assets to be written off in the first year, there may be an incentive to overstate their value. If substantial customs duties are applied to such imports, this abuse is unlikely-but one hopes that econo- mies emerging from socialism will employ only low and uniform tariffs. International Issues The United States might not allow foreign tax credits for the business SAT because interest expense is not deductible.9 If that happens, a country adopting the SAT might forfeit the opportunity to have part of its taxes paid by the U.S. government. A noncreditable SAT might discourage investment by U.S. multi- national corporations. This issue is easily overstated. First, creditability is an issue only for repatri- ated earnings (see Hartman 1985). Second, a tax with a METR of zero, even combined with the U.S. tax on repatriated earnings, is unlikely to discourage investment. Third, since the 1986 U.S. tax reform, many U.S. multinationals have paid more foreign taxes than they can credit against U.S. liability, and so 232 The World Bank Research Observer, vol. 7, no. 2 (July 1992) they generally would not be able to take a credit even if the tax were a credit- able income tax.10 There is, however, more to the story than this. First, not all U.S. firms have excess foreign tax credits. Second, similar questions arise for other countries employing worldwide taxation and offering foreign tax credits for source- based income taxes. Third, if the United States raised its corporate tax rates, any country adopting a noncreditable tax could be in a worse position than if it had a creditable tax. Finally, creditability may have symbolic importance. In this it may resemble treaties on double taxation. Developing countries do not appear to benefit, on balance, from such tax treaties with industrial countries, but the treaties, by providing evidence of a stable investment envi- ronment, may bestow intangible benefits. Much the same may be said of cred- itability. Transition Issues The shift from an income tax to the SAT would pose difficult transition prob- lems for a market economy. Introducing such a tax would involve "radical re- forms of the present tax system and... cannot be implemented without a great upheaval" (Sinn 1987, pp. 348-49; see also Aaron and Galper 1985; McLure and others 1990). For countries emerging from socialism, the transition would be much easier. Take the issue of how to treat the undepreciated basis of assets and interest on outstanding debt to avoid windfall gains and losses. Shifting to nondeduct- ibility of interest is not a major change for socialist economies in transition, because some interest is already nondeductible under enterprise taxation based on "balance sheet" profits. In addition, interest rates have been exceptionally low. Because there has been no deduction for depreciation, there is no issue of how rapidly to allow the remaining basis to be written off. If the SAT were introduced before privatization, windfalls would be experi- enced by state-owned enterprises, rather than by individuals and privately owned companies. Windfalls would also be less troubling under some forms of privatization-for example, if the SAT were introduced just after all citizens re- ceived equal shares (or vouchers) in all enterprises and no private investment had yet occurred. The longer that introduction of a new tax system is delayed, the more transition issues resemble those in the West. The status quo is not a viable option for socialist economies in transition; they must abandon their present tax systems. Transition problems will occur whether the shift is to an SAT or an income tax. Transition to an income tax may be more problematical, because of timing issues and inflation adjust- ment. At high rates of inflation, most nominal interest payments should be disallowed as deductions or excluded from income; disregarding all such pay- ments involves only a minor difference. The administrative and compliance Charles E. McLure, Jr. 233 effort needed for a tax on real economic income is far greater than that need- ed for the SAT. Any problems in introducing the SAT pale beside the transition effects of moving from a command economy, with state ownership and controlled prices, to a market economy with private ownership and market-determined prices. The transition period offers unparalleled freedom to choose the most appro- priate tax system, relatively unencumbered by past choices. "No One Does It" No country has wanted to be the first to introduce the SAT. This argument is less telling in countries emerging from socialism: no country has ever made the transition from a command economy to free markets, either. Still, these countries do not want to be guinea pigs in a fiscal experiment. In assessing the thesis presented here that the SAT is appropriate for socialist economies in tran- sition, it is important to realize that for most persons the two systems would be virtually indistinguishable; only the treatment of income from business and capital is different. The primary potential costs of choosing the SAT are the up- front revenue loss of expensing and the subsequent transition problems of shift- ing to an income tax and making interest taxable and deductible (and the re- lated problems of changing the tax treatment of dividends). Debtors would not object to deductibility of interest on existing indebtedness (or for dividends); the main problem would be revenue loss. Creditors would object to taxation of interest income from existing debt, and shareholders would object to taxa- tion of dividends on existing stock. It would be undesirable to allow interest deductions without taxing interest income. Since the marginal tax rates of most borrowers are probably higher than those of most lenders, borrowers and lend- ers would probably voluntarily restructure debt. Relation to the Value Added Tax Although the SAT is commonly (as here) seen as a substitute for the income taxes, it resembles a subtraction-method VAT (see McLure forthcoming). But because the SAT can accommodate personal exemptions and graduated rates on labor income, it has distributional advantages not shared by VATs. This raises the question: should a country have both a VAT and the SAT? One can imagine substituting the SAT for the VAT, individual and company income taxes, and payroll taxes. The answer depends in part on whether the country aspires to join the European Community (EC). If so, it must adopt an EC-type VAT, and its only choice is between the SAT and an income tax. For other countries the question is more complicated. Administrative resources would be saved by concentrating them on one tax, but at the risk of putting "all the eggs in one basket." 11 234 The World Bank Research Observer, vol. 7, no.2 (July 1992) Concluding Remarks Countries emerging from socialism must quickly implement tax systems that will finance the proper functions of government in a noninflationary manner. Yet they are ill-prepared to cope with the intricacies of a standard income tax. They lack the requisite accounting practices, tax administration, and experi- ence with compliance. Tax policy must address these limitations. Indeed, ad- ministrative considerations are almost as important as economic effects in the choice of a tax system for reforming socialist economies. The SAT encourages saving and investment, is economically neutral, and avoids many administrative problems of an income tax, especially timing issues and the need for inflation adjustment. The SAT is not a panacea. But it deserves consideration. Notes Charles E. McLure, Jr., is a Senior Fellow of the Hoover Institution, Stanford University. This article is based on McLure (forthcoming), prepared for the Socialist Economies Unit of the World Bank's Country Economics Department. The author has benefited from comments by Richard Bird, Cheryl Gray, Janos Kornai, Stephen Langlois, John Litwack, and George Zodrow (whom he credits with teaching him most of what he knows about the SAT). Other helpful comments came from participants in a seminar organized by the Fiscal Affairs Department of the Interna- tional Monetary Fund, including Milka Casanegra, Ved Ghandi, Leif Muten, Abdul Rahman, Vito Tanzi, and Christine Wallich. 1. This section draws on McLure 1988 and 1989. 2. This section draws on earlier work with George Zodrow reported in Zodrow and McLure (forthcoming), McLure and others (1990), and McLure and Zodrow (1990). See also Hall and Rabushka (1983 and 198S); Bradford (1986); U.S. Department of the Treasury (1977); Mintz and Seade (1991). 3. This differs from the treatment of interest and debt under the expenditure tax (the "R plus F" base of Institute for Fiscal Studies 1978), under which the proceeds of borrowing and interest income are included in the tax base, and repayment of debt and interest expense are deductible. 4. If inflation is extremely high it may be desirable to adjust nominal amounts of transactions occurring within a given year to make them comparable; this possibility is ignored. The payment of interest on losses carried forward is more problematic if the inflation rate-and thus the in- terest rate-is high and variable. 5. McLure (forthcoming) discusses other economic effects of the SAT, including incentives for investment, tax holidays, investment hunger, and countercyclical effects. 6. Progressivity can be altered by adjusting the rate structure. A more highly graduated rate structure would produce disincentives that could retard economic development. Moreover, grad- uated rates in the SAT are not as effective as those in the income tax in dealing with concentration in the distribution of income from capital, since the METR is zero under the SAT. 7. Sinn (1987, p. 349) offers the following pessimistic observation: "A reform that leaves the rentier's return untaxed cannot be made palatable to any of the world's parliaments." 8. SAT rules should not be adopted for financial accounting. They are not appropriate for judg- ing business performance. 9. McIntyre (1990) argues that the "R plus F" type of consumption-based tax (Institute for Fiscal Studies 1978) would not be creditable and should not be. His arguments appear to be ap- plicable to the SAT (see also Sunley 1988; Kay 1988). McLure (1990) argues that the SAT should be creditable. The SAT is less burdensome than a tax on real net economic income, as indicated Charles E. McLure, Jr. 235 by METRs of zero under the SAT and the statutory tax rate under the income tax. Also, the tax base is likely to be smaller under the SAT. The SAT has economic and administrative advantages; it is not a gross receipts tax levied to soak up foreign tax credits. For extraordinary profits, the portion of the income tax base that is common to the base of the SAT, the two taxes are identical. The case for creditability should be presented carefully, to avoid dooming the SAT to the ranks of noncreditable taxes. 10. In effect, the United States allows credit only to the extent that foreign taxes do not exceed the amount implied by application of the average U.S. tax rate of the company to foreign-source income. Under the "overall limitation" on the credit, foreign-source income and foreign taxes for all countries (but not all types of income and income taxes) are lumped together in calculating the limitation. For a thorough discussion see Slemrod (1991). Goodspeed and Frisch (1989) esti- mate that the fraction of foreign-source income subject to excess credits would increase from 50 percent to 78 percent (from 32 percent to 82 percent in manufacturing) as a result of the 1986 changes in U.S. tax rates. 11. Border tax adjustments (compensating taxes on imports and tax rebates on exports) might not be allowed for the SAT as VAT, on the grounds that the SAT is not an indirect tax, the only type for which border tax adjustments are allowed under the General Agreement on Tariffs and Trade. References Aaron, Henry J., and Harvey Galper. 1985. Assessing Tax Reform. Washington, D.C.: Brookings Institution. Boskin, Michael J., and Charles E. McLure, Jr., eds. 1990. World Tax Reform: Case Studies of Developed and Developing Countries. Published for the International Center for Economic Growth. San Francisco: ics Press. Bradford, David E 1986. Untangling the Income Tax. Cambridge, Mass.: Harvard University Press. Goodspeed, Timothy J., and Daniel J. Frisch. 1989. U.S. Tax Policy and the Overseas Activities of U.S. Multinational Corporations: A Quantitative Assessment. Washington, D.C.: Office of Tax Analysis, U.S. Department of the Treasury. Gray, Cheryl W. 1991. "Tax Systems in the Reforming Socialist Economies of Europe." Commu- nist Economies and Economic Transformation 3, no. 1: 63-79. Hall, Robert E., and Alvin Rabushka. 1983. Low Tax, Simple Tax, Flat Tax. New York: McGraw-Hill. . 1985. The Flat Tax. Stanford, Calif.: Hoover Institution Press. Hartman, David G. 1985. "Tax Policy and Foreign Direct Investment." Journal of Public Eco- nomics 26, no. 1 (February): 107-21. Institute for Fiscal Studies. 1978. The Structure and Reform of Direct Taxation: Report of a Committee Chaired by Professor J. E. Meade. London: George Allen & Unwin. Kay, J. A. 1988. "The Implementation of a Cash Flow Corporation Tax." Paper presented at World Bank conference, The Feasibility and Implementation of Cash-Flow Company Taxa- tion, Washington, D.C. King, Mervyn A., and Don Fullerton. 1984. The Taxation of Income from Capital. Chicago: University of Chicago Press. Kornai, Janos. 1990. The Road to a Free Economy-Shifting from a Socialist System: The Case of Hungary. New York: W W. Norton. McIntyre, Mike. 1990. "Editor's Notebook." Tax Notes International (July): 691-93. McLure, Charles E., Jr. 1979. Must Corporate Income Be Taxed Twice? Washington, D.C.: Brookings Institution. 236 The World Bank Research Observer, vol. 7, no. 2 (July 1992) . 1988. "The 1986 Act: Tax Reform's Finest Hour or Death Throes of the Income Tax?" National Tax Journal 41, no. 3 (September): 303-15. . 1989. "Lessons for LDCs of U.S. Income Tax Reform." In Malcolm Gillis, ed., Tax Re- form in Developing Countries. Durham, N.C.: Duke University Press. . 1990. "International Aspects of Tax Policy for the 21st Century." American Journal of Tax Policy 8, no. 2 (Fall): 167-85. . 1991. "Tax Policy for Bulgaria," Bulletin of the International Bureau for Fiscal Docu- mentation 45, no. 5 (May): 235-47. . Forthcoming. "A Consumption-Based Direct Tax for Countries in Transition from So- cialism." Wayne Law Review. McLure, Charles E., Jr., Jack Mutti, Victor Thuronyi, and George R. Zodrow. 1990. The Tax- ation of Income from Business and Capital in Colombia. Durham, N.C.: Duke University Press. McLure, Charles E., Jr., and George R. Zodrow. 1990. "Administrative Advantages of the Indi- vidual Tax Prepayment Approach to the Direct Taxation of Consumption." In Manfred Rose, ed., Heidelberg Congress on Taxing Consumption. Proceedings of the International Congress on Taxing Consumption, Heidelberg, June 28-30, 1989: 335-82. Berlin: Springer Verlag. Mintz, Jack M., and Jesus Seade. 1991. "Cash Flow or Income? The Choice of Base for Com- pany Taxation." World Bank Research Observer 6, no. 2 (July): 177-90. Sinn, Hans-Werner. 1987. Capital Income Taxation and Resource Allocation. Amsterdam: North-Holland Press. Slemrod, Joel. 1991. "Developing Country Tax Policy Toward Foreign Direct Investment in the Light of Recent International Tax Developments." In Anwar Shah, ed., Fiscal Incentives for Investment in Developing Countries. World Bank, Country Economics Department, Washington, D.C. Processed. Sunley, Emil. 1988. "The Treatment of Companies under Cash Flow Taxes: Some Administrative, Transition, and International Issues." Paper presented at World Bank conference, The Feasi- bility and Implementation of Cash-Flow Company Taxation, June 20, Washington, D.C. U.S. Treasury Department. 1977. Blueprints for Basic Tax Reform. Washington, D.C.: Govern- ment Printing Office. (Also available as D. F. Bradford and the U.S. Treasury Tax Policy Staff. 1984. Blueprints for Basic Tax Reform. Arlington, Va.: Tax Analysts.) Zodrow, George R., and Charles E. McLure, Jr. Forthcoming. "Implementing Direct Taxes on Consumption in Developing Countries." Tax Law Review. Charles E. McLure, Jr. 237