chapter 1 Why a Value-Added Tax?

Alan Tait
Published Date:
June 1988
  • ShareShare
Show Summary Details

The latest innovation is the value-added tax. Its emergence in France illustrates the process by which a sort of continuing ferment of improvisation now and then gives rise to an invention of the first order.

—Carl S. Shoup, “Taxation in France,” National Tax Journal,

Vol. 8, December 1955, p. 328

The rise of the value-added tax (vat) is an unparalleled tax phenomenon. The history of taxation reveals no other tax that has swept the world in some thirty years, from theory to practice, and has carried along with it academics who were once dismissive and countries that once rejected it. It is no longer a tax associated solely with the European Community (EC). Every continent now uses the vat, and each year sees new countries introducing it. Various similes come to mind; vat may be thought of as the Mata Hari of the tax world—many are tempted, many succumb, some tremble on the brink, while others leave only to return, eventually the attraction appears irresistible. This book has a modest intent. It tries to illustrate how different countries have tackled the problems of vat that are common to most. It debates the major issues and discusses preferred solutions. The book is divided into three main sections: the practice and problems of vat in terms of vat’s structure (Part I), effects (Part II), and administration, including taxpayer compliance (Part III).

In this opening chapter there are three main parts. First, a brief account of the different ways a tax can be levied on value added. Second, an examination of the reasons why countries decide to switch to a vat. Finally, a most important consideration, why some countries have decided (for the time being?) not to adopt a vat.

VAT in Theory

Four Basic Forms1

Value added is the value that a producer (whether a manufacturer, distributor, advertising agent, hairdresser, farmer, race horse trainer, or circus owner) adds to his raw materials or purchases (other than labor) before selling the new or improved product or service. That is, the inputs (the raw materials, transport, rent, advertising, and so on) are bought, people are paid wages to work on these inputs and, when the final good or service is sold, some profit is left. So value added can be looked at from the additive side (wages plus profits) or from the subtractive side (output minus inputs).

If we wish to levy a tax rate (t) on this value added, there are four basic forms that can produce an identical result:

  • (1) t (wages + profits): the additive-direct or accounts method;

  • (2) t (wages) + t (profits): the additive-indirect method, so called because value added itself is not calculated but only the tax liability on the components of value added;

  • (3) t (output – input): the subtractive-direct (also an accounts) method, sometimes called the business transfer tax; and

  • (4) t (output) – t (input): the subtractive-indirect (the invoice or credit) method and the original EC model.

Direct Versus Indirect Methods

Given that there are four possible ways of levying a vat, why has only one (method 4, above) been popular? Most taxes are levied by first calculating the tax base (income, sales, wealth, property values, and so on) and then applying a tax rate to that value. The same method might be thought sensible for a vat. The value added should be calculated directly (methods 1 or 3, above) and the tax rate applied. In practice, the method used (number 4) never actually calculates the value added; instead, the tax rate is applied to a component of value added (output and inputs) and the resultant tax liabilities are subtracted to get the final net tax payable. This is sometimes called the “indirect” way to assess the tax on value added.

Why should an indirect method of tax calculation be used for vat when the alternative seems so much more straightforward? There are four principal reasons. The most important consideration is that the invoice method (number 4) attaches the tax liability to the transaction, making it legally and technically far superior to other forms. As will be explained in later chapters, the invoice becomes the crucial evidence for the transaction and for the tax liability.

Second, as discussed in Chapter 13, the invoice method creates a good audit trail. Experience in countries such as Benin and Mauritania that use method 3—what in French is called the base sur base method—suggests that without the invoice for each transaction problems emerge, first, in ensuring that inputs are deducted only when tax is paid and, second, when inputs exceed taxable sales. This same method has been described, in Canada, as not leaving a good audit trail and as practically eroding the revenue base despite “the Calvinistic nature of the Canadian populace.”2

Third, to use methods 1 and 2, which are accounts based, profits need to be identified. As company accounts do not usually divide sales by different product categories coinciding with different sales tax rates, and as they certainly never divide inputs by differential tax liabilities, it is clear at once that the only vat that could be levied on an additive basis would be a single rate vat. If a multiple rate vat is wanted, it rules out using methods 1 and 2.

Finally, the easiest way to calculate a vat, using the subtractive method, appears to be the calculation of the value added (output minus input) and then to apply the tax rate to that figure (method 3). In practice, companies do not find it convenient to calculate their value added in this way month by month, as purchases, sales, and inventories can fluctuate greatly. Firms may have to carry stocks that change, according to the type of production, the seasonality of trade, or anticipated interruptions of supplies. Again, this procedure is only practical using a single rate. In fact, calculating the direct value added is easiest through the trader’s annual accounts, and so this method of deriving a vat (in addition to methods 1 and 2) is also an “accounts method.”

Thus, to date, method 4, the invoice or credit method, is the only practical one. The tax liability can be calculated week by week, monthly, quarterly, or annually. It is the method that allows the most up-to-date assessments and also allows more than a single rate to be used.

Capital Purchases

This brings us to another problem skated over in the initial presentation. How do we treat substantial purchases of long-lasting inputs (that is, capital goods)? According to the methods already mentioned, capital purchases would be inputs and would be deducted from any sales. This, of course, can cause huge fluctuations in tax liability as the purchase of, say, a new factory could occur in one month, and lead to negative value added in many of the succeeding months. Alternatively, using the additive method, profits are usually calculated after allowing for only a portion of the cost of a capital purchase (depreciation). Rarely do income tax authorities allow traders to expense their capital inputs (that is, treat each purchase as an immediate expense so that buying a factory becomes the same as buying an automobile or buying a meal); instead, elaborate rules are designed to allow different assets to be depreciated over different lengths of time (for example, machines over 7 years and buildings over 20 years). To make a vat calculated under the full consumption base of the subtraction method (where all capital purchases are offset at once against sales) exactly the same as a vat on the additive base, a different calculation of profits would have to be adopted. Depreciation would be abolished and all capital purchases would be offset at once in the accounts, thus making profits much smaller in the early years of capital purchases and much larger in later years, when the usual depreciation would not be deducted. Clearly, the profits shown for income tax purposes in the “profit and loss accounts” would differ hugely from those calculated for vat.

This is not to say that one way is right and the other wrong—just that they produce very different results.

To the Retail Sale?

Although vat is usually thought of as applied to all stages of production including the retail sale, this is not necessarily the case. Table 1-1 shows nine countries that do not apply the vat at the retail stage (though two, Grenada and India, are not really using a vat although the name is implied in their legislation). Two countries (Morocco and Peru) apply the vat through the wholesale stage and the others only to the manufacturing level. It should be mentioned that some countries (for example, Kenya) employ a manufacturers sales tax that allows credit for tax paid on inputs and is, therefore, practically a single-stage vat.

Table 1-1.Countries That Do Not Apply a VAT Through the Retail Sale
VAT Applied at Level of
BrazilFederal VATState VAT to retail sale
Côte d’IvoireX
Grenada1XDifferential rates on imports and domestic goods
India1XMixture of ad valorem and specific excise rates
Source: See text.

Not really a VAT; see section on “Tax Evolution and Efficiency,” below.

Source: See text.

Not really a VAT; see section on “Tax Evolution and Efficiency,” below.

It is clear that all these less-than-complete vats create problems. All involve a much smaller tax base than one which includes retail sales and, therefore, their tax rates must be higher to yield an equivalent revenue. A vat to the wholesale level must define a wholesale price because traders often combine manufacturing and wholesale activities as well as wholesale and retailing activities; this leads to a complex set of rules or regulations defining “up-lifts” from factory gate prices or establishing standard discounts on retail prices. A vat through the wholesale stage should only be considered as a temporary interim arrangement on the way to extending the vat fully to the retail stage. It is doubtful whether the inefficiencies for both taxpayer and administration ever make it worthwhile other than as a temporary arrangement.

The vat on manufacturing and importation is more common. It allows a developing economy to levy a buoyant tax on an ad valorem principle and accustom its traders to a credit system. Frequently, the small manufacturers are exempt and, de facto, the vat is a tax on imports and on large, well-organized industry, especially multinationals. After a few years’ experience, the manufacturing vat can be extended to the retail level. While this is a more attractive option than the vat to the wholesale level, it involves an even smaller tax base. However, experience gained from the more limited base and the use of credits allows the vat to be extended to the sale of all goods and services. (Argentina, Bolivia, and Korea used a manufacturing sales tax with a credit mechanism before moving to a complete vat.) Of course, there are still problems. In Morocco, where there are some vertically integrated firms, the vat, although formally levied on manufacturers and wholesalers, can extend right through to the retail stage; a provision allows that an enterprise selling to a connected enterprise (that is, wholesaler to retailer) has to pay vat on behalf of the retailer. This provision is only included to catch cases of abuse, and, where a genuine internal transfer price can be proved, it is accepted. Nevertheless, this sort of provision exemplifies the difficulties faced by systems that do not apply vat through the retail stage.

Prices Inclusive or Exclusive of Tax?

The final possible variation in vat is whether or not the tax rate is levied on a price inclusive or exclusive of the tax liability. A 10 percent vat, on a price exclusive of vat, is clear to the consumer. However, it does mean letting the purchaser know both the price before vat is applied and the amount of tax that must be paid. Alternatively, if the tax is quoted tax inclusive, then a rate of only 9.1 percent would be needed to generate the same revenue (9.1 percent of 110 yields 10). In theory, it makes no difference which method is used; in practice, though France originally used it, only Finland and Sweden have persistently employed the tax-inclusive base.

On another pricing issue, U.S. commentators often claim incorrectly that the vat conceals the tax burden from the consumer. Consumers in Europe can see goods priced with or without vat; if without, they are then exposed to the shock of vat added at the cash register. If anything, this is a much greater “tax shock” than the usual tobacco, gasoline, or liquor excise common in the United States. Any sales tax can be “concealed” from the taxpayer; this is not a criticism restricted only to vat.

The Choice

So, in theory, we have the choice of adopting one of four basic forms of vat, allowing capital purchases to be fully expensed or depreciated over time, and quoting the rate on a price inclusive or exclusive of tax. In practice, the choice has been much more limited.

As noted already, the methods of deriving vat from company accounts basically require a single rate tax, or a business or trade that deals in a particularly homogeneous commodity; for example, the hotel trade (which is dealt with later). Most countries do not wish to build in this lack of flexibility to their vat in the initial legislation (although Japan has suggested this method); so this leaves only one practical option, the subtractive-indirect method. Indeed, this is the type of vat adopted by almost all countries to date.

However, while theory points us toward this conclusion, it was not theory that persuaded countries to adopt the invoice or credit method of calculating vat. France, the first country to use a vat, did not sketch the pros and cons of the different ways of levying vat, as we have just done, and then implement it. In practice, why do countries adopt a vat?

Why Adopt the VAT?

Countries introduce a vat because they are dissatisfied with their existing tax structure. This dissatisfaction falls broadly into one, or possibly all, of four categories: (1) the existing sales taxes are unsatisfactory; (2) a customs union requires discriminatory border taxes to be abolished; (3) a reduction in other taxation is sought; or (4) the evolution of the tax system has not kept pace with the development of the economy.

Unsatisfactory Sales Taxes

The Cascade Tax

The simplest sales tax is one that takes a straightforward percentage of all business turnover. Because tax on tax occurs as a taxed product passes from manufacturer to wholesaler to retailer, this has become known as a cascade tax. The defects of this type of tax are well known3 and have caused most countries using it to switch.4 As Table 1-2 shows, most of the early users of vat switched from various forms of cascade taxes. Indeed, it was the disadvantages of the cascade element that persuaded the French, first, to allow a credit for the tax content of purchases of raw materials against the sales tax liability and, second, to allow a credit for the tax content of capital purchases. In this way, the vat using the invoice method developed from ways to mitigate the disadvantages of the original cascade tax. Indeed, Finland, for example, levies a sales tax at an effective rate of 19.05 percent using a credit mechanism, but does not allow the purchase value of fixed assets, fuel, and other goods consumed in the business to be deducted. In this way, Finland is still at a stage preparatory to a full vat.

Table 1-2.Summary Showing Substitution of VAT for Other Taxes and Proposed Effect on Revenue


Sales Taxes

Mainly Replaced1


On Revenue


Tax Changes
ArgentinaJan. 1975Wholesale sales and provincial cascade turnover taxEqual yieldProvincial tax changes
AustriaJan. 1973Cascade wholesaleEqual yieldLower income tax
BelgiumJan. 1971Cascade wholesaleEqual yield
BoliviaNov. 1973Multistage ring systemEqual yield or increaseIndividuals can offset vat against a 10 percent tax on gross income: a 1 percent cascade turnover tax retained
Brazil2Jan. 1967Cascade tax on sales and consignmentsEqual yieldChanges in federal-state revenue shares; separate tax on services
ChileMar. 1975Cascade turnover, manufacturers tax, and special luxury taxIncreaseGasoline, income, and property taxes raised
ColombiaJan. 1965No previous sales taxIncreaseIncome, property, and capital gains taxes changed
Costa RicaJan. 1975Multistage ring systemIncreaseIncreased excises
Côte d’IvoireJan. 1960Manufacturers vatEqual yield
DenmarkJuly. 1967WholesaleIncreaseIncreased excises
Dominican Rep.Nov. 1983No previous sales taxIncreaseReduce reliance on customs duties
EcuadorAug. 1970Turnover taxes on mining and manufacturingIncreaseMining taxes reduced
FranceJan. 1968An earlier and less sophisticated vatEqual yieldTax exemptions abolished and income tax adjustments
Germany, Fed. Rep. ofJan. 1968Cascade retailEqual yield
GrenadaApr. 1986Stamp duties and taxes on servicesEqual yieldCorporate and income taxes abolished and corporate cash flow tax introduced
GreeceJan. 1987Turnover tax, stamp duties, and special import levyEqual yield“Luxury tax” retained
GuatemalaAug. 1983Stamp duty on sales, services, and imports
HaitiNov. 1982Seventy-nine excisesEqual yieldReplaced commissions, levies, and excises
HondurasJan. 1976Single-stage ring systemIncrease
HungaryJan. 1988Production and turnover taxesIncreaseMassive tax and subsidy restructuring including corporate and personal income taxes
India3Apr. 1985
IndonesiaApr. 1985Manufacturers ring system with eight ratesEqual yieldReform of the income tax
IrelandNov. 1972Wholesale and retail salesEqual yieldSome tariff reductions
IsraelJuly. 1976Various salesIncrease
ItalyJan. 1973General and local government salesEqual yield
Japan4Apr. 1989No sales taxVAT would increase revenue but offset by corporate and personal tax reductions; net revenue neutralReduction in corporate and personal income taxes and changes in excises
KoreaJuly. 1977Eight sales taxes representing 40 percent of revenueEqual yieldChanged excises
LuxembourgJan. 1970Cascade wholesaleEqual yield
MadagascarJan. 1969Cascade productionIncrease
MexicoJan. 1980Cascade production and revoked 18 selective sales taxesEqual yield or increaseLower border vat of 6 percent
MoroccoJan. 1962Cascade productionEqual yieldChange in corporate and production taxes
NetherlandsJan. 1969Cascade wholesaleEqual yieldLower income tax
New ZealandMay. 1986Wholesale taxYield extra revenueChanges in personal and corporate income taxation and fiscal incentives and low income credit introduced
NicaraguaAug. 1978Multistage ring systemEqual yieldReduced customs duties
NigerJan. 1986Cascade manufacturersYield extra revenueReplace existing taxes on services
NorwayJan. 1970Sales taxes on 65 percent of consumptionLossReduced income and property taxes
PanamaJan. 1976No sales taxIncreaseStamp taxes reduced and increased excises
PeruJan. 1973Cascade production and stamp taxIncrease
PhilippinesJan. 1988Eight sales and stamp taxesIncrease
Poland41989Restructure income taxes
PortugalJan. 1986Single-stage wholesaleEqual yieldAbolition of stamp duties and minor taxes
SenegalMar. 1961Manufacturers vatEqual yield
South Africa4Apr. 1989Retail sales taxEqual yieldSubstantial review of direct and indirect taxes
SpainJan. 1986Cascade production tax and 20 other sales taxes
SwedenJan. 1969Retail sales tax and capital goods taxEqual yield1 percent payroll tax to offset lost revenue
Taiwan Province of ChinaApr. 1986Cascade retail tax and stamp dutyEqual yieldSeparate sales tax on financial services, night clubs, and small businesses
Thailand4Jan. 1989Business turnover taxEqual yield
TurkeyJan. 1985Eight production taxes and other duties on goods and servicesEqual yieldAd hoc export tax rebates phased out and individuals allowed to deduct vat as a withholding offset against a 10 percent tax on gross income; offset diminishes as income rises
United KingdomApr. 1973Multirate wholesaleLossSelective employment tax removed
UruguayJan. 1968Manufacturers single-stage tax and a cascade turnover taxEqual yield
Source: Various country reports and see text.

This column is as accurate as a brief summary can be: “cascade production tax” refers to a cascade tax on business turnover restricted to the production stage; “cascade wholesale tax” extends the turnover tax to include the wholesale stage; “cascade retail tax” extends the turnover tax to include the retail stage; “manufanucturers,”“wholesale,” or “retail” taxes are single-stage taxes, some operated on a ring system, others on a credit system.

Note Brazil introduced a federal vat on interstate transactions and a state vat on intrastate sales (ICM); however, the Federal Government determines the tax base and rates.

Not actually a vat but a federal excise (with both ad valorem and specific rates) with credit for purchases (but not for capital goods).


Source: Various country reports and see text.

This column is as accurate as a brief summary can be: “cascade production tax” refers to a cascade tax on business turnover restricted to the production stage; “cascade wholesale tax” extends the turnover tax to include the wholesale stage; “cascade retail tax” extends the turnover tax to include the retail stage; “manufanucturers,”“wholesale,” or “retail” taxes are single-stage taxes, some operated on a ring system, others on a credit system.

Note Brazil introduced a federal vat on interstate transactions and a state vat on intrastate sales (ICM); however, the Federal Government determines the tax base and rates.

Not actually a vat but a federal excise (with both ad valorem and specific rates) with credit for purchases (but not for capital goods).


Manufacturer and Wholesale Taxes

Not as simple as the cascade, but much fairer, is a single-stage tax levied on manufacturers, wholesalers, or retailers. However, some undesirable cascading can be found even in countries that use or have used single-stage sales taxes (Australia, Austria, Belgium, Canada, Luxembourg, Mexico, Morocco, the Netherlands, New Zealand, and the United Kingdom). Cascading occurs whenever taxable goods are produced using taxed inputs. The problem is more frequently found when the point of impact of the tax is far from the retail stage; for instance, as in a manufacturers or importers tax, since the likelihood of a manufacturer buying his inputs from another manufacturer or from an importer is much greater than buying them from a retailer. But even retail sales taxes, by including in their definition of taxable sales the sales of certain kinds of producer goods that can also be used as consumer goods, can produce cascading.5

The difficulties of operating a manufacturers sales tax are well exemplified by the Canadian example. Since in a manufacturers sales tax the tax liability can differ sharply according to the source of the inputs and the amount of integration between the manufacturing, wholesale, and retail stages, Canadian businesses reduced their tax liability by setting up related, but separate, distribution companies. This meant the markup of the distribution company did not enter the tax base. To combat the loss of revenue, the administration devised a set of rules (effective July 1988) that create complex uncertainties for traders.6 Treating the symptoms rather than the disease is a classic case of the trail that leads countries away from unsatisfactory sales taxes to the vat.

Cascading can also occur when a variety of different sales taxes are used. Some countries become dissatisfied not only with the complexity of administration but also with the complex and multiple relationship between traders and government when many taxes are used. In Korea, eight indirect taxes (a business turnover tax, a commodity tax, and taxes on textiles, petroleum, gas and electricity, travel, admissions and entertainment, and food) were replaced by the vat and a supplementary “special consumption tax” (more or less an excise). In this way, the Government sought to simplify tax administration as each previous tax had its own rate structure (from 0.5 to 300 percent), a different tax base, and administrative procedures.7 The tax content of exports, because of cascading, was unknown.

One way to avoid the worst8 cascading is to allow credit for some purchases. Indeed, this has been a common way of moving (unconsciously in the case of some countries) toward a full vat. However, sometimes countries find (for example, South Africa) that the revenue cost of giving up taxing, say, capital goods, makes it difficult to give credit, or at least, to give full credit immediately. As a result, the credit for tax is gradually and sporadically extended and the system moves toward a full vat.

Other mechanisms to avoid cascading are in use. One that is fairly common under manufacturers or importers sales taxes is to allow manufacturers registered as taxpayers to acquire tax free, locally produced goods or imported goods used as inputs. This is the so-called ring system (Bolivia, Costa Rica, Honduras, Indonesia, and Nicaragua before the vat; see Table 1-2) where sales between registered traders (within the ring) are tax free, but sales to unregistered taxpayers outside the ring—such as retail sales—are taxed. The main problem with this system is the weight placed on the registration number. Traders are meant to verify that they sell to or buy from only appropriately registered traders free of tax; however, such is the value of this privilege that many traders invent or abuse registration numbers and evasion can reach epidemic proportions.

The best solution, even for a manufacturers or importers sales tax, is probably a credit system similar to that in use for vat even if it does not fully compensate for all tax paid on inputs (for example, because of exemptions). One additional advantage of this system is that at least a part of the tax is payable when the inputs are purchased, and any subsequent action by the manufacturer to evade taxation results in a loss of only part of the tax. Tax administrators always prefer cash in the bank that taxpayers must claim back; a bird in the bush is difficult to see, never mind catching. Another advantage is that it accustoms traders to working with a tax system that depends on the invoice and on a credit for tax paid, and is halfway home to a vat. Indeed, some Central American countries (Costa Rica, Honduras, and Nicaragua), as in France originally, have adopted a vat almost as though by accident through moving from a ring system to a credit mechanism.

Some countries, such as Australia and New Zealand, have used single-stage wholesale taxes and found them unsatisfactory. These wholesale taxes were introduced many years ago (in 1933 for New Zealand) and exhibited many signs of their age. Numerous exemptions and multiple rates complicate the structure and erode the base of single-stage wholesale taxes. Politicians try to classify goods according to their “luxury” nature, and also attempt to exempt some goods as potential business inputs. Exemptions become based on the nature of the purchaser or the end use of the goods instead of according to the nature of the goods.9 Retail prices are distorted because different retail markups are not taxed and it pays to adjust transfer prices to maximize the untaxed markups. Retailers carry tax-paid stocks, which affect their liquidity and which limit the ability of the wholesale tax to respond quickly as a demand management tool to changes in tax rates. Exports frequently must absorb some of the domestic wholesale tax, as is the case in Canada. This situation places exporters at a disadvantage or leads to a depreciated exchange rate. Finally, services cannot be included under the wholesale tax. For all these reasons, countries find the single-stage tax on intermediate goods unsatisfactory.

It is clear that with any sales tax, the higher the tax rate, the more incentive there is for taxpayers to evade the tax. Assuming a given need for revenue on the part of the government, the form of sales tax that provides the broadest possible base is the one that requires the lowest rate. From this point of view, either a retail sales tax or a vat carried through the retail level has a comparative advantage over sales taxes with more limited bases.

Retail Sales Tax

For countries with a sophisticated tax administration and good taxpayer compliance a retail sales tax is an attractive option. The 45 U.S. state sales taxes, at rates of 4–8.25 percent, are efficient sources of revenue, raising about 20 to 60 percent of tax revenue in Vermont and Washington, respectively, with an average of 33 percent. However, even under favorable conditions, problems begin to accumulate as the tax rate rises. At 5 percent, the incentive to evade tax is probably not worth the penalties of prosecution; at 10 percent, evasion is more attractive, and at 15–20 percent, becomes extremely tempting. Only Iceland, Norway, South Africa, Sweden, and Zimbabwe have operated retail sales taxes at rates over 10 percent. Both Norway (with a retail sales tax rate of 13.64 percent) and Sweden (11.1 percent) decided to switch to a vat even though, as countries outside the EC, they did not have to adopt it. South Africa is dissatisfied with its 12 percent retail sales tax and is exploring ways to change to a vat. Iceland now has a retail sales tax of 25 percent, and the continuous concerns about evasion have resulted in legislation to introduce a vat.

Zimbabwe levies retail sales taxes at 15 percent and 18 percent, yielding about a quarter of total tax revenue. This may be the most successful example of a retail sales tax levied at high rates, though there is increasing worry about evasion.10 Nevertheless, in a relatively simple economy, there seems to have been a widespread acceptance of the tax though this may have been due, in part, to the major portion of revenue secured from a few major retailers.

Switzerland uses what can be called a retail sales tax (although basically it is a wholesale tax) and has twice rejected a vat. Finland uses a manufacturers tax and a retail sales tax at 6 percent and 4 percent, respectively. However, “there has long been consideration of shift to a general value-added tax.”11 Finally, Paraguay levies a retail sales tax on the larger retailers and on sales to smaller retailers by manufacturers and wholesalers at rates ranging from 4 percent to 14 percent. The vat has not been adopted in Paraguay because of fear of the additional paperwork required.

There is substantial discussion about the retail sales tax versus the vat,12 but the main problems of the former seem to be the following:

  • The higher the rate, the more collection weight is put upon the weakest link in the chain—the retailer, especially numerous small retailers.

  • All the revenue is at risk. It has been suggested that this is also true of vat if the retailer successfully claims all his credit on purchases, but clearly it is more difficult to do so under the accounting requirements of vat.

  • The audit and invoice trail is poorer than under a vat, especially for services.

  • There have to be troublesome “end-use exemptions.”

  • Revenue is not secured at the easiest stage, that is, at the time of importation, and this can be crucial for many developing countries.

In other words, a single point retail sales tax is efficient at relatively low rates, but is increasingly difficult to administer as rates rise. This, of course, can be seen as an advantage by those who wish to restrain the capacity of the government to raise revenue through broadened sales taxes (a common refrain in the U.S. Congress). By association, a vat that partly circumvents this disadvantage can be caricatured as a license for government profligacy.

Interestingly enough, Norway is the only country that, having changed from a retail sales tax to a vat, has considered seriously the possibility of switching back. Many held that the vat compliance costs were high and the opportunity for fraud greater. However, a Treasury paper13 argued that the vat fell more certainly on final sales than the retail sales tax, it was more secure (19 percent of the registered firms made 90 percent of the sales), it improved accounting standards, and the possible advantages of the retail sales tax in no way compensated for the transitional costs involved in moving back to a single-stage tax.

Others have also pointed out that a vat may be less open to lobbying influences than the retail sales tax; a successful lobbying effort on the retail sales tax removes the tax from the entire industry, but lobbying under a vat is more diffuse as there is the issue of zero rating versus exemption (see Chapter 3) and the likelihood that upstream traders will still be left with tax liabilities on inputs if the good is only exempted but not zero rated.

Special Needs of Countries in Regional Economic Groupings

The vat was the choice of the member countries of the EC as the best way to promote neutrality and uniformity of the tax burden and to provide incentives for increased productivity and industrialization. The recommendation of the EC Fiscal and Financial Committee that all member countries shift to the vat form was formulated in 1962.14 The change for the original members was finally completed in 1973, when Italy implemented the tax.15 However, other recommendations regarding uniformity of rates and eventual adoption of the origin principle for taxation of trade within the EC have not been put into effect and indeed the EC is now firmly on the path of the destination principle (see Chapter 8 for a discussion of this issue).

Most of the countries in the Latin American regional groups—the Latin American Integration Association (LAIA, formerly known as the Latin American Free Trade Association or Lafta) and the Andean Pact—have adopted value-added taxes. Brazil also has a federal manufacturers vat which, like the Colombian manufacturers vat, features a credit system to avoid cascading.16 Although the Latin American countries in the LAIA have a long way to go before reaching the stage of tax harmonization that can be found in the EC today, it appears likely that the vat will eventually become the harmonized form of general sales taxation in the region.17

The vat can be applied either on the basis of origin or on the basis of destination—which then allows replacement of the destination principle by the origin principle when a significant degree of harmonization has been achieved. However, this is one of those textbook arguments that are always presented, yet are wildly out of touch with reality; the possibility of any free trade area adopting the origin principle of taxation must be decades away.18 Indeed, the application of the origin principle with a vat could be an administrative nightmare.19

Of course, none of these issues need preoccupy countries that do not intend to be part of a customs union. However, the vat is the only common sales tax in a customs union, other than a retail sales tax, that fulfills the obligations for tax neutrality on traded goods and services under the General Agreement on Tariffs and Trade (GATT).

A Reduction of Other Taxation or to Simply Increase Revenue?

Some countries look to a vat not only to replace existing sales taxes but also to increase revenue (see Chile and Denmark in Table 1-2). However, some have viewed it as a new source of revenue which will enable other taxes to be reduced or abolished. Much of the debate in the United States about vat has revolved around the possibility of replacing the corporate profits tax, reducing the rate of individual income tax, permitting property tax relief, financing social security, reducing the payroll tax, or, more generally, reducing the public sector borrowing requirement.

Until 1986, no country had used the vat to reduce the corporate profits tax. The early British review20 rejected vat as a substitute for the corporate profits tax, largely on grounds that the substitution of a tax that would be passed forward for one that was not must increase prices. The increased prices would decrease the competitive position of exports, and the balance of payments on current account would deteriorate. Increased wage demands to maintain real wages would increase costs and this would further decrease the international competitive position of exports. Numerous arguments can be mounted against this exposition and it can be shown that with different assumptions, prices need not rise and profits and investment could increase.21

In the United States, it is claimed that the substitution of vat, which will be fully rebated on exports, for the corporate income tax, which is not rebatable, would improve the competitiveness of U.S. industry and improve the balance of trade. This somewhat simplistic argument can be rejected.22 Nevertheless, proposals are made for such a substitution. However, the fall in the effective rate of corporate taxation in recent years has meant this substitution is no longer such a live issue.23

New Zealand also amended its corporate income tax in connection with vat, but this was to increase the rate from 45 percent to 48 percent to align the top rate with the new top rate for personal income taxation; so the change is better seen associated with personal taxes than with sales taxes.

Increasing the revenue collected from the indirect tax base, when replacing existing sales taxes by a vat, allows adjustments in direct taxes. Although Denmark increased revenue when the vat was introduced in 1967, this permitted adjustments in income tax to offset, partly, any consequent change in prices. Similarly, it can be argued that top marginal rates of income tax could be reduced and the lost revenue replaced by vat (although the switch would be regressive). A more current debate is on using vat revenue to reduce the burden of payroll taxes to finance social security. It has been suggested that businesses in Argentina saw an advantage in a vat, which would be rebated on exports and which might be more difficult to collect, replacing part of the payroll tax, which was not rebated, was easily assessed, and was collected quite efficiently.

In general, much of this desire to use vat to replace other taxes flows from dissatisfaction with an increased reliance on direct taxes. As Table 1-3 shows, every member country of the Organization for Economic Cooperation and Development (OECD) collected a smaller proportion of its revenue from indirect taxation in 1975 than in 1965. The fall for some was as much as 42 percent (Spain) or 30 percent (Belgium). The required increase in direct taxation to offset the falling importance of indirect taxes was, as a proportion of gross domestic product (gdp), as high as 100 percent (Turkey) or 71 percent (Switzerland). Faced with such problems, some countries (Denmark, Luxembourg, and the United Kingdom) used the vat during 1975–84 to stem the tide, halting the increase in direct taxes and transferring some of the burden to the vat. Other countries (Italy, France, and Norway) were less successful in containing the increase in direct taxes, and some (Belgium, Ireland, and Sweden), as a percentage of gdp, combined substantial increases in both direct and indirect taxes.

The most interesting countries, however, are some of those that have not introduced a vat. In 1965–82, Australia increased its reliance on direct taxes from 65 percent to 68 percent of total revenue; Canada, from 59 percent to 67 percent; Japan, from 74 percent to 85 percent; and Switzerland, from 69 percent to 81 percent. In each of these countries, the introduction of vat has, to date, been debated but rejected or delayed. This may suggest that were vat to be introduced, increasing reliance on direct taxes could be halted and switched to indirect taxes.

Table 1-3.Taxes on Goods and Services as Percentage of Total Taxation in OECD Countries
Percentage Change
Germany, Fed. Rep. of3332272727–180–18
New Zealand2827242227–1412–4
United Kingdom3329252930–2420–9
United States2219181718–180–18
Source: Organization for Economic Cooperation and Development, Revenue Statistics of OECD Member Countries, 1965–1985 (Paris, 1986), p. 94.
Source: Organization for Economic Cooperation and Development, Revenue Statistics of OECD Member Countries, 1965–1985 (Paris, 1986), p. 94.

However, in practice, as the examples of Belgium, Ireland, and Sweden show, such an outcome is by no means automatic. Stopping the rise in the direct tax burden requires more than the introduction of a VAT. Strong political decisions are needed to prevent “bracket creep,” to contain the use of proportional payroll taxes for social security, and to resist the temptation to use a buoyant VAT revenue as simply another way to increase expenditure.

Certainly, VAT as a buoyant revenue source, closely linked to increases in consumption, has become a crucial part of overall revenue for all countries using it. Indeed, many countries find that their initial revenue from the first year of vat turns out substantially higher than forecasts based on past sales tax bases or on national income accounts. Indonesia’s manufacturing level vat, “in the first year of operation … managed to raise revenues about 45 percent over target … even in a country that has a reputation of having a weak tax administration.”24 As Tables 1-4 and 1-5 show, vat frequently contributes 15–25 percent of central government tax revenue and 5–10 percent of gdp. The experience of countries using vat has been all one way—upward (except for Costa Rica where the vat was reduced from 10 percent to 8 percent within a few months of its introduction in response to public protests; also, a consolidation of rates in Peru left the standard rate of vat at 18 percent instead of 20 percent originally). In all other countries, rates of vat in 1988 were at least the same as they were at the introduction of vat and were usually higher.

Tax Evolution and Efficiency

Countries are supposed to move from simple to complex tax structures and from systems that distort allocation to those that are more neutral. The easiest point of taxation for countries with a simple economic structure is at importation. Customs duties are one of the oldest forms of tax, visible and not too difficult to administer. Excises, usually with specific rates and relying on physical controls, follow. However, as trade becomes more complex, countries are pushed, as far as indirect taxes are concerned, toward introducing a truly general sales tax for its revenue buoyancy and lack of distortion.

This need not always occur. For instance, countries in the Middle East do not use general sales taxes. Numerous excises are levied and the intent may be to add commodities one by one to move toward a general sales tax, but progress has been slow. Socialist regimes in Eastern Europe use turnover taxes, sometimes employing over a thousand different tax rates. Other countries recognize that a multiplicity of ad hoc sales taxes at different rates is inefficient both administratively and economically, and they move to vat to consolidate and modernize their tax structure (for example, Chile, Haiti, Hungary, Indonesia, Korea, and Mexico).

Usually, countries find the demand for government revenue to be such that a broad-based sales tax is desirable. Relying on selective sales taxes with narrow bases becomes increasingly distortionary. While some move through the stages of credit and ring systems, there is a tide in the affairs of tax men which leads, if not to greatness, at least to generalness. As mentioned earlier, the vat has become a fashionable tax, accounting for important shares of revenue and gdp (Tables 1-4 and 1-5). While fashion may not be the best way to develop a tax system, it is, undoubtedly, a powerful influence.

The more vat systems there are, the more likely it is that a country, developing its indirect tax systems, will adopt a vat. There is also some common sense in this. The more examples there are to follow, the less likelihood of mistakes. Legislation and regulations can be adopted to suit the particular contingencies of a country, but it is better to have half a dozen alternative laws and experiences to start from than none at all. The visible success of vat in many countries in generating buoyant revenues is a selling point for other authorities looking for a modern revenue base.

Indeed, a remarkable recent development can be ascribed partly to this “fashionable” growth of vat. The first socialist economic system to adopt vat has been Hungary in 1988. The objective of the tax reform was to “diminish the tax burden on the business sector, to make it more uniform, and to simplify taxation and budgetary relationships.”25 This is an enormously ambitious, complex tax and price reform involving the introduction of a uniform corporate income tax, a personal income tax, and a vat, along with reductions in corporate subsidies and reform of social security financing. It is interesting that when the authorities were examining all the possible options for a completely new sales tax, they eventually settled on the invoice or credit method of vat; “basically, the Hungarian system will be in line with international practice.”26 From recent pronouncements, it sounds as though the Polish authorities may be following the same path arguing that their present tax system no longer fully meets the needs of the economy and that the system needs to evolve to improve efficiency.27

Another reform that involves a vat to finance a gradual lowering of tax rates throughout the tax system is that recommended in the Report of the Margo Commission in South Africa.28 However, the form of vat suggested is unique. The tax is named the comprehensive business tax, which is an origin-based, direct-additive accounts vat (method 1 discussed above). The tax base would include salaries and wages, interest, royalties, rent, profit, and depreciation; gross investment would be subtracted but exports would be taxed. This base is, of course, equal to sales less purchases.29 The Margo Commission Report emphasizes that the comprehensive business tax should be regarded as an income tax but is, however, a vat. Moreover, it is a vat that does not exempt exports (basically because of the numerous unchecked borders in Southern Africa that make border sales tax adjustments difficult). If enacted, this would have been a remarkable, and closely watched, experiment. However, the Government has rejected this recommendation and now proposes to introduce a normal accounts-based vat in 1989.

Table 1-4.VAT as Percentage of Total Tax Revenue, 1975–86
Costa Rica10.
Côte d’Ivoire10.611.7
Germany, Fed. Rep. of13.112.912.413.
United Kingdom10.09.89.510.411.916.514.315.315.916.817.917.6
Source: International Monetary Fund, Government Finance Statistics Yearbook, Vol. 11 (Washington, 1987).
Source: International Monetary Fund, Government Finance Statistics Yearbook, Vol. 11 (Washington, 1987).
Table 1-5.VAT as Percentage of GDP, 1975–86
Côte d’Ivoire2.22.5
Germany, Fed. Rep. of3.
United Kingdom3.
Source: International Monetary Fund, Government Finance Statistics Yearbook, Vol. 11 (Washington, 1987).
Source: International Monetary Fund, Government Finance Statistics Yearbook, Vol. 11 (Washington, 1987).

It should be pointed out that there are at least two so-called vats that are not vats at all. Grenada imposed a “VAT” in 1986 at the manufacturing level at 20 percent. However, the 20 percent rate was applied to imports and to only 40 percent of the value of domestic sales. What this means in practice is that the tax is highly discriminatory on imports (20 percent), compared with domestic manufacturing (20 percent on 40 percent is equivalent to only 8 percent on total sales). This is really a tariff of 20 percent and a local manufacturing tax of 8 percent.

The so-called Indian modvat (modified value-added tax) is not a vat at all, but rather a form of modified excise duty.30 It is a manufacturing excise tax (levied at both ad valorem and specific rates) with credit allowed for excise duty and customs in a limited number of industries. It may eliminate the cascading effect of multipoint excise levies. Introduction of a full vat in India would seem to present numerous administrative and constitutional difficulties, including the vexed question of union-state relations.31

A VAT—But Not Just Yet

Given the worldwide movement to adopt vat, the examples of countries that have considered a vat but have decided against it are particularly interesting.

Japan is an example of a country that has held out against the vat from the original Shoup proposal in 1953 until 1986 when, finally, proposals for a vat were approved by the ruling party.32 However, this proposal was also rejected by Parliament. Australia and New Zealand both flirted with vat in the early 1980s; Australia rejected vat and New Zealand introduced it (but the opposition party promised to repeal it). Greece was committed to change to a vat on entering the EC in 1981 as a full member, but successive prevarications delayed the introduction until January 1987. However, no matter how sporadic the moves toward vat, the direction of change seems irresistible. Yet this is what makes the cases of those countries that have resisted this movement so interesting. The examples of the United States, Canada, Australia, Japan, and Iceland are all informative.

United States

The first point is that the United States is fairly satisfied with its current state and local retail sales taxes. Suggestions for a vat at the federal level immediately run into the practicality of allowing the federal government to appropriate a general sales tax, and the problem of adding or “piggybacking” the state sales taxes on to a federal vat. This is the overwhelming consideration for evaluating the possibility of a U.S. vat (see Chapter 8).

In addition to the problems with the retail sales tax mentioned earlier, the United States has further difficulties. The experience of the states with retail sales taxes has been at relatively low rates (the median rate for 1985 was 4.6 percent) and frequently on a limited base (29 states do not tax food and 25 exempt most services); such administrative experience does not necessarily carry across to ensure an efficiently run vat at, say, double the rate on all goods and services and levied on all traders.

For the United States to adopt a vat, there has to be some reason other than dissatisfaction with the existing structure. Americans have been interested in vat for the last twenty years,33 but to give the flavor of recent debates, we can select five reasons given for suggesting the introduction of a vat—shown in Table 1-6, along with the commentators and the dates of debate.

In the late 1970s, Congressman Ullman, in H.R. 5665, the (unsuccessful) Tax Restructuring Act of 1979, proposed a vat to replace part of the payroll tax ($52 billion), the personal income tax ($50 billion), and the corporate income tax ($28 billion). In the event, Congressman Ullman was defeated in the November 1980 election (perhaps not entirely disassociated with his advocacy of vat), but his proposals continued to attract attention. But by 1981, Charles McLure was writing, “if it is decided that a federal sales tax is needed, it would probably be better to adopt a federal retail sales tax than to impose a new and unfamiliar form of sales tax, the vat.”34

Table 1-6.Summary of Some Reasons for Advocating a U.S. Value-Added Tax






Income Tax


Income Tax


Charles E. McLure, Jr., “The Tax Restructuring Act of 1979: Time for an American Value-Added Tax?” Public Policy (Cambridge, Massachusetts), Vol. 28 (Summer 1980), pp. 301–22.

Charles O. Galvin, “It’s VAT Time Again,” Tax Notes, Tax Analysts (Arlington, Virginia), Vol. 21 (October 24, 1983), p. 280.

Richard W. Lindholm, A New Federal Tax System (New York: Praeger, 1984).

Senator William V. Roth, Jr., The Business Transfer Tax Act of 1985 (S. 1102, 99th Congress, 1st Session, May 8, 1985) and “The Roth Reforms” (speech to the National Press Club, Washington, February 20, 1986).

See discussion in Charls E. Walker and Mark A.Bloomheld, eds., The Consumption Tax: A Better Alternative? papers presented at a conference sponsored by the American Council for Capital Formation—Center for Policy Research (Cambridge, Massachusetts: Ballinger Publishing Company, 1987).

Charles E. McLure, Jr., The Value-Added Tax: Key to Deficit Reduction? (Washington: American Enterprise Institute, 1987).

Charles E. McLure, Jr., “The Tax Restructuring Act of 1979: Time for an American Value-Added Tax?” Public Policy (Cambridge, Massachusetts), Vol. 28 (Summer 1980), pp. 301–22.

Charles O. Galvin, “It’s VAT Time Again,” Tax Notes, Tax Analysts (Arlington, Virginia), Vol. 21 (October 24, 1983), p. 280.

Richard W. Lindholm, A New Federal Tax System (New York: Praeger, 1984).

Senator William V. Roth, Jr., The Business Transfer Tax Act of 1985 (S. 1102, 99th Congress, 1st Session, May 8, 1985) and “The Roth Reforms” (speech to the National Press Club, Washington, February 20, 1986).

See discussion in Charls E. Walker and Mark A.Bloomheld, eds., The Consumption Tax: A Better Alternative? papers presented at a conference sponsored by the American Council for Capital Formation—Center for Policy Research (Cambridge, Massachusetts: Ballinger Publishing Company, 1987).

Charles E. McLure, Jr., The Value-Added Tax: Key to Deficit Reduction? (Washington: American Enterprise Institute, 1987).

In 1983, during the debates on the need to increase defense spending, Charles Galvin35 suggested that a way to “market” the vat would be to “appeal” to the American people that a 10 percent vat on all purchases would finance defense and permit the budget to be balanced. This, of course, is really just a different way of saying that the vat is needed to increase net revenue. He also suggested that if it were considered that tying vat to defense might prove a political liability, then why not tie it to financing social security? Again, of course, enabling the budget to move toward a balance.

A much more radical proposal was put forward in 1984 by Richard Lindholm.36 The corporate and personal income taxes and the estate taxes were to be replaced by a vat at 15 percent and a net worth tax at 2 percent; the vat would be the principal generator of current revenue and the net worth tax would contribute a better equity component than the present direct taxes. As far as vat is concerned, this proposal, of course, concentrates on the nondistorting revenue buoyancy of vat.

The 1985 proposal by Senator Roth37 was for a business transfer tax (BTT) to be levied, using the direct-subtractive method. That is, all traders with turnover exceeding $10 million, except retailers, were to subtract inputs from outputs leaving value added (basically wages and profits) to be taxed. This vat would be rebated on exports as is the usual invoice-subtractive vat but, in addition, it was proposed to allow the vat liability as a credit against the social security tax (fica); this suggestion seriously undermines the validity of the export rebate under GATT as it transforms the vat from a sales tax to a direct tax. Such a direct-subtractive vat is only practical using a single rate applied universally—“Given the likelihood that pressures for exclusions and multiple rates would not be successfully withstood, it seems inadvisable to adopt a naive subtraction-method tax such as the btt.”38

Charls Walker39 polled 22 former high-level government economic policymakers and found broad support to reduce the budget deficit. Two thirds supported a tax increase, and nearly all of these preferred a consumption tax. This could be a retail sales tax or a vat but its primary aim would be to generate a net revenue increase and reduce government borrowing.

Finally, the title of Charles McLure’s book, The Value-Added Tax: Key to Deficit Reduction? puts the emphasis squarely on the need for revenue. “Interest in the … vat … will increase as the Gramm-Rud-man-Hollings targets for deficit reduction become increasingly difficult to achieve through budget cuts.”40

The opposition to a U.S. vat is summed up in five points.41

  • Liberals oppose the vat on grounds of regressivity.

  • Conservatives fear the vat as a “money machine.”

  • Both liberals and conservatives worry that a vat would be inflationary.

  • State and local officials are concerned about a vat’s intrusion into their traditional preserve for raising revenue.

  • Both federal and state officials fear that the vat would be an administrative nightmare.

Against these feelings the principal thrust behind the U.S. proposals to introduce a vat is the need to supplement federal revenue.42 Such a tax change could also allow social security to be financed, encourage savings, and permit direct taxes to be restructured; but the main problem is that the Federal Government does not have access to the buoyant revenue of a general sales tax. (It is estimated that a general 5 percent vat would raise $70 billion in 1988 and a vat exempting food, housing, and medical care would raise $40 billion.)43 What form of general sales tax is another debate, and one well illustrated by the Australian debate (the U.S. federal-state issue is mirrored in the Canadian debate, below).


The Australian wholesale tax is admitted to be administratively complex, inequitable, and inefficient.44 The choice for Australia to generate more revenue and achieve more flexibility to reduce personal marginal income tax rates boiled down to introducing a retail sales tax or a vat. In the event, both proved unacceptable and it was decided to extend the existing wholesale tax and introduce a separate tax on services.

The reasons Australia decided against a vat, despite the advocacy of such bodies as the Business Council of Australia, included the following:

  • The self-policing properties of vat had been overstated.

  • The administrative costs could be “unacceptably high.”

  • The incentive to evade tax was as high under a vat as under a retail sales tax.

  • Overstated claims for tax credit was as significant a potential loophole as were falsified invoices.

  • Compliance costs were higher under a vat.45

  • Trade union concerns about regressivity (a 12.5 percent vat was expected to increase prices by about 6.5 percent and there were doubts about the promised compensation for welfare benefits).

However, reading between the lines, a possibly overwhelming consideration was the much greater “lead-time” that the vat involved. The Australian Commissioner of Taxation estimated that the additional lead-time needed to introduce a vat rather than a retail sales tax could be of the order of 12 months.46 In a country where there is a constitutional parliamentary term of three years, the time taken to examine, legislate, introduce, and get over the initial problems of a controversial new tax can become a crucial consideration. Indeed, it is interesting that political considerations, rather than economic, may have been the deciding factor in the delays in changing to vat in countries as diverse as Australia, Canada, Cyprus, Greece, and Japan.


Rather like Australia, Canada relies on an unsatisfactory manufacturers tax that is widely thought to involve distorting cascade tax elements in exports. A 1984 survey indicated that the average effective tax rate for domestic goods was 33 percent higher than the tax on imports competing with domestic products. Successive amendments to this tax have turned it into a set of hybrid manufacturers or wholesalers tax; several broad categories of goods are taxed at the wholesale level (for example, cosmetics, automobiles, televisions and audio goods, and household chemicals). As in Australia, Canada has toyed with the idea of a vat.47 However, unlike Australia the main difficulty has involved the federal-state issue. All but one of the 11 Canadian provinces have substantial independent retail sales taxes that they are reluctant to give up.48 This suggested that a direct-subtractive vat (the business transfer tax) based on company accounts might be more suitable because its effects on prices are not explicit and a retail sales tax could still be added to it. While it is true such a method might enable vat revenues to be allocated to the states where the value added originated, this would depend on the universality of the tax and, of course, on the use of a single rate. It would also be a unique experiment watched most carefully by other major federal systems (Brazil, India, Nigeria, and the United States). The difficulty of reaching agreement with the provinces (quite apart from all the usual concerns about regressivity and price effects) has probably been the major influence restraining Canada’s adoption of the vat.


Japan has been on the brink of introducing a vat before. In 1980, the Finance Ministry tried to introduce a vat-style “consumption tax” and was repulsed. The 1986/87 vat at 5 percent was not put forward as a revenue-enhancing measure, but to replace revenues lost because of reducing the top individual income tax rate from 70 percent to 50 percent and the corporate income tax rate from 53 percent to 50 percent.

There seem to be three main problems with the proposal. First, the principal opposition to the vat appears to come from a widespread suspicion that a vat would give the Finance Ministry a revenue buoyant tax that could greatly expand the scope of government. At present, Japan relies mainly on excises and stamp duties for its indirect tax revenue. As reported, many think that “the Finance Ministry has long been in search of additional revenue sources, particularly the vat. . . . In the long run, we might have a European-type large-spending government.”49

Second, although the proposed vat had a high exemption limit of ¥ 80 million (US$627,000) and would have exempted most of the six million individual businesses, many of these small traders supply large taxable companies and would have been unable to supply a vat registration number allowing their customers to claim the vat as a deduction. This, it was claimed, would put them at a disadvantage. There seems to have been a feeling among small traders that you were damned if you did and damned if you didn’t register for vat.

Finally, the bill included the obligation of taxable enterprises to seek code numbers from their tax offices. Although designed to prevent fraud, it was perceived that the new system could easily be applied to income and corporation taxes. Such a registration system was resented and helped sentiment against the tax that introduced it.

This controversial vat, as part of a wider tax reform package, was unable to obtain Parliamentary approval. However, another suggestion to introduce vat may be debated during the summer of 1988.


For the last ten years, Iceland has considered introducing a vat.50 It is the only country that has levied a retail sales tax sustained at over 13 percent for many years. Introduced in 1960 at 3 percent, it mounted to 11 percent by 1970, and was supplemented by an “emergency tax” and “oil tax” in 1973/74 that increased it to 13 percent. By 1986 the base retail sales tax was 20 percent, the oil tax 1.5 percent, and an “additional sales tax” 3.5 percent, making 25 percent in all. Such a rate does cause considerable concern as collection lies entirely in the hands of retailers, since registered traders import free of tax. Evasion is suspected, and exemptions of foods and many services exclude about 40 percent of total private final consumption. Administration is, of course, fairly straightforward in a country of 210,000 where 50 percent of the population lives in the capital city. Although the Federation of Icelandic Industries has supported the vat, its introduction has been delayed for two reasons. First, trade unions have opposed it because of anticipated unfavorable effects on prices and, second, the administrative problems of monitoring the refunds has caused doubts. However, the attractions of collecting the major proportion of the sales tax at importation (typical of a small island economy) overrode many doubts, and the proposal is to introduce the vat at 24 percent, although delayed again from January 1988 to January 1989.


This brief review shows that the reasons why countries have not introduced vat boil down to:

  • Fear about regressivity.

  • Anticipated high administrative costs, especially to monitor refunds.

  • Buoyant revenue permitting larger government expenditures.

  • Potential evasion.

  • Compliance costs.

  • Effect on prices.

  • Incompatibility of vat with a traditionally strong state sales tax structure.

Each of these concerns is dealt with separately in later chapters. Overall, we have the picture of a general sales tax that has exploded from its final form in France in the late 1960s to become the dominant sales tax of the world. The theory of the vat has been looked at many times; it is worth looking at the practice, at the problems encountered, and how the examination of such problems may improve practice.

There are numerous ways to present the theory of vat. This is the bare bones of the debate. The literature is replete with somewhat tedious arithmetical examples of how the vat works and it is not intended to repeat them here. For a clear version, see United States, Department of the Treasury (1984, Vol. 3, Chap. 2) and for a more comprehensive treatment, see McLure (1987b, Chap. 3).

Jenkins (1986, pp. 11–12).

Cumulative and unknowable tax liabilities, different tax liabilities depending on the degree of industrial vertical integration, difficulty in assessing the amount of cascade tax to be rebated on exports or imposed on imports, and so on; see Cnossen (1987c).

As a rough rule of thumb, it is estimated that the effective rate of a cascade tax to the retail stage is approximately two and a half times the nominal rate: thus, a turnover tax of 4 percent is equivalent to a retail sales tax of 10 percent.

See Shoup (1973, pp. 220–21).

See discussion in “TPI Country Survey—Canada: Sales Tax Reform,” Tax Planning International Review (1987).

Choi (1984, pp. 1–6).

But not all—see Hemming and Kay (1981, pp. 80–81).

In a classic example, New Zealand exempted motorcycles used for farm work but not for ordinary road use; the distinction is not easy in practice. On the difficulties of administering a wholesale sales tax, see Cnossen (1983, pp. 316–23).

Due (1986b, p. 243).

See the discussion in Due and Brems (1986, pp. 11–12).

Members joining the EC later had to introduce a vat; Ireland (1972), the United Kingdom (1973), Portugal (1986), Spain (1986), and Greece (1987).

Oddly enough, the names of the vat differ: impuesto al valor agregado (Argentina, Mexico, and Uruguay); impuesto sobre las ventas (Colombia, Costa Rica, and Honduras); impuesto a las transacciones mercantiles y prestación de servicios (Ecuador); impuesto a la transferencia de bienes corporales muebles con crédito fiscal (Panama); and impuesto a los bienes y servicios (Peru). Perhaps, as in New Zealand (goods and services tax), countries thought that by calling a rose by another name, it might smell more sweetly and its thorns would be ignored?

Consideration of these aspects in more detail can be found in European Community, Commission of the European Economic Community (1963, pp. 102–103).

United Kingdom, Report of the Committee on Turnover Taxation (1964, especially Chap. 8).

Tait (1972, Chap. 7).

See evidence of Richard A. Musgrave in United States, Congress (1972, pp. 131–32) and Tait (1972, pp. 103–108).

But C. Lowell Harriss testified to the Ways and Means Committee of the U.S. Congress on September 27, 1984 that the corporate tax should be eliminated and a vat substituted. See United States, Congress (1984, pp. 330–37).

Jenkins (1986, p. 11).

Lukács (1987, p. 447).

Lukács (1987, p. 449).

Reported in the Wall Street Journal (New York), December 29, 1987, p. 10.

South Africa, Report of the Commission of Inquiry into the Tax Structure of the Republic of South Africa (1987).

South Africa, Report of the Commission of Inquiry into the Tax Structure of the Republic of South Africa (1987, p. 347).

Chakravarty (1986, pp. 41–43).

Despite this long gestation period, one commentator said, “The bill was apparently hastily drafted.” See “Japan: Sales Tax,” World Tax Report (1987, p. 14).

For the earlier period, see the discussion in McLure (1974, pp. 96–103).

McLure (1981, pp. 156–57); see also McLure (1980a).

Senator William V. Roth, Jr., The Business Transfer Tax Act of 1985 (S. 1102) and “The Roth Reforms.” speech to the National Press Club, Washington, February 20, 1986.

McLure (1987b, p. 68).

See discussion in Walker and Bloomfield (1987).

McLure (1987b, p. 1).

See “Commentary,” in Bloomfield (1987, p. 174).

See also United States, General Accounting Office (1986).

“Congressional Reports: Estimated Revenue Effects of Options to Raise Revenue,” Tax Notes (1987, p. 88).

Australia, Reform of the Australian Tax System: Draft White Paper, Appendix 13–B, “Choice Between vat and bbct” (1985, pp. 129–32).

Australia, Reform of the Australian Tax System: Draft White Paper, Appendix 13–B. 17 (1985, p. 131).

See a comment by Gordon Pitts, “Value Added Tax Examined,” Financial Post (Toronto), October 27, 1984, p. 15.

See the interesting discussion in Gillis (1986).

Makino (1986, p. 4).

Due (1986b, pp. 240–41).

    Other Resources Citing This Publication