chapter 10 Effects of VAT on Prices

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

O chestnut tree, great-rooted blossomer,

Are you the leaf, the blossom, or the bole?

O body swayed to music, O brightening glance,

How can we know the dancer from the dance?

—W.B. Yeats, “Among School Children”

The effects of vat, like those of most other taxes, are difficult to assess empirically. It is almost impossible to link, unequivocally, the change in vat to the change in prices, efficiency, investment, employment, or whatever. There are usually too many variables to relate one change uniquely to another. Indeed, in some countries the introduction of the vat coincided with the massive upheavals following the sharp increases in oil prices in 1972/73 and 1976 (Ireland, 1972; Austria, Bolivia, Italy, and the United Kingdom, 1973; and Honduras, Israel, and Peru, 1976), and the effects of the vat pall besides those of the oil price changes. In some instances, we can outline what economic analysis might suggest and in a few examine the evidence of the actual outturn.

In all countries politicians and the public anticipate the impact of vat on retail prices with trepidation. Studies are conducted to estimate the possible outcomes, and governments, with varying degrees of commitment, try to protect the public from price increases. A characteristic of all introductions of the vat is uncertainty. Traders, particularly small businessmen, are uncertain what effect the vat will have on their liquidity and on the costs of compliance; the public hear bits and pieces of information, and newspapers sometimes enhance the air of crisis by “scare” stories. In these circumstances, traders might well attempt to widen margins as a contingency against uncertainiy, and the public may be persuaded to accept higher prices because speculation has suggested they should expect them.1 Let us examine first, what common sense suggests should happen to prices when a vat is introduced and when vat rates change, then, how governments actually coped with the problems in real life, and finally review such empirical evidence as there is.

Price Change Forecasts

The first point is whether or not the vat is intended to increase revenue or simply replace revenue lost from other taxes reduced or replaced. Table 10-1 shows that, in the majority of cases, the vat introduction was designed to be an equal yield replacement for other taxes. In this case, the analysis must consider the effects of reducing the other tax or taxes. In an extreme example, replacing eight taxes in Korea by vat could imply a complex mixture of changes in factor prices and producer prices at manufacturing and wholesale levels. Similarly, in Colombia, the changeover to a full vat was combined with changes in income, property, capital gains taxes, and the abolition of many tax incentives. The problem in most countries is to find the data in sufficient detail to allow any ex ante estimates. Korea was unusual in taking great care to prepare a sample of several hundred businesses to assess the extent to which each of the taxes to be replaced by vat were reflected in prices and, by implication, the price reduction or increase that might be expected when the taxes were removed and vat imposed. This was a major exercise but it paid off in the detailed substantiated position the Government was able to take when the vat was introduced (see below).

Table 10-1.Examples of Effects on Prices of Introduction of VAT

Price Change


Taxes Mainly



on Revenue

(In percent)

to VAT3

(In percent)


Tax Changes
Any Other

Concurrent Changes

ArgentinaJan. 1975Wholesale sales taxEqual yield37.2MinorProvincial tax changesUtility rates increased; devaluationYes, but relaxed
AustriaJan. 1973Cascade wholesale taxEqual yield2.4NilLower income taxesStrict credit controlYes
BelgiumJan. 1971Cascade wholesale taxEqual yield2.6NilIncreased wagesYes
BoliviaOct. 1973Multistage ring systemEqual yield or increase9.5NilNew luxury tax rates; increased excisesNo
BrazilJan. 1967State sales and municipal industrial taxesEqual yield15.8NilNo
ChileMar. 1975Cascade turnover taxIncrease146.7MinorTaxes on gasoline, income, and property raisedUtility rates increased and rent controls relaxedNo
ColombiaJan. 1975Simpler vatIncrease12.9NilIncome, property, and capital gains taxes changedMany incentives abolishedNo
Costa RicaJan. 1975Multistage ring systemIncrease4NilIncreased excisesNo
Côte d’IvoireJan. 1960Manufacturers vatEqual yieldNo
DenmarkJuly 1967Wholesale taxIncrease8.05.0Lower income taxIncreased transfers and increased wagesNo
EcuadorJuly 1970vat was a new tax; some turnover taxes on mining and manufacturing replacedIncrease8.7(7.1)Mining taxes reducedDevaluatonNo
FranceJan. 1968Simpler vatEqual yield2.11.0Tax exemptions abolished and income tax adjustmentsIncreased wagesYes, after vat introduced
Germany, Fed. Rep. ofJan. 1968Cascade retail taxEqual yield1.50.6Monitored
HondurasJan. 1976Multistage ring systemIncrease1.0NilRapidly expanded creditNo
IndonesiaApr. 1985Production taxEqual yield3.5NilNo
IrelandNov. 1972Wholesale and retail sales taxEqual yield5.5NilSome tariff reductionsMonitored
IsraelJuly 1976Various sales taxesIncrease17.9(9.0)Increased wages and tax allowancesNo
ItalyJan. 1973Central and local government sales taxEqual yield6.3NilIncreased wagesNo
KoreaJuly 1977Eight sales taxes representing 40 percent of revenueEqual yield4.1MinorChanged excisesProposed vat rate of 13 percent reduced to 10 percentYes
LuxembourgJan. 1970Cascade wholesale taxEqual yield3.5NilNo
MadagascarJan. 1969Cascade production taxIncrease3.2NilNo
MexicoJan. 1980Cascade production taxesEqual yield or increaseLower border vat of 6 percentNo
MoroccoJan. 1962Cascade production taxEqual yield2.4NilChange in corporate and production taxesNo
NetherlandsJan. 1969Cascade wholesale taxEqual yield5.21.5Lower income taxIncreased wagesYes
New ZealandOct. 1986Wholesale taxIncrease8.96.5Reduced income taxNo
NicaraguaJan 1975Multistage ring systemEqual yield4NilReduced customs dutiesNo
NorwayJan. 1970Sales taxes on 65 percent of consumptionLoss7.85.8Reduced income and properly taxesIncreased transfers and wagesYes
PanamaMar. 1977Wholesale taxIncrease5.0(3.1)Stamp taxes reduced and increased excisesIncreased utility ratesNo
PeruJuly 1976Cascade production taxIncrease27.1(13.5)No
PortugalJan. 1986Single-stage wholesaleEqual yield10.0(2.0)No
SenegalMar. 1961Manufacturers vatEqual yieldNo
SpainJan. 1986Cascade sales taxEqual yield2.82.0Major administrativeMonitoring
SwedenJan. 1969Retail sales taxEqual yield1.6Nil1 percent payroll tax to offset lost revenueNo
TurkeyJan. 1985Nine production taxes and dutiesEqual yield40.0(10.0)Wage earners rebate and vouchersNo
United KingdomApr. 1973Multirate wholesale taxLoss4.90.7Selective employment tax removedMonitoring
UruguayJan. 1968Manufacturers, wholesale, and retail taxesEqual yield66.3(53.0)No
Sources: IMF data and various country reports.

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: “manufacturers,” “wholesale,” or “retail” taxes are single-stage taxes—some operated on a ring system, others on a credit system.

The consumer price index for the quarter before vat is deducted from that for the quarter after vat and the difference expressed as a percentage increase.

Figures in parentheses represent broad estimates based on examination of the data: other figures are based on external and contemporary commentaries.

Monthly data unavailable for part of the period covered.

Sources: IMF data and various country reports.

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: “manufacturers,” “wholesale,” or “retail” taxes are single-stage taxes—some operated on a ring system, others on a credit system.

The consumer price index for the quarter before vat is deducted from that for the quarter after vat and the difference expressed as a percentage increase.

Figures in parentheses represent broad estimates based on examination of the data: other figures are based on external and contemporary commentaries.

Monthly data unavailable for part of the period covered.

Even if reasonable data are available, some major theoretical issues must be faced. What is to be the assumption on shifting? The vat is expected to be passed forward fully both because the legislation usually clearly assumes that this will happen and because, as a general sales tax, there is likely to be a general awareness on the part of traders that all will be affected similarly. Sometimes traders will take advantage of the confusion to raise prices “exorbitantly” and even to levy vat when they are not liable (for example, taxi drivers in Portugal).2 Naturally, differential rates will cause changes in relative prices. All traders will bear tax and compliance costs and wish to recoup them.

Moreover, the public is likely to be aware of a major tax change and gloomily expect the worst. But every business could pass along to its customers the full amount of its vat liability only if the “enactment of the vat were accompanied by repeal of the law of supply and demand.”3 Of course, the vat need not be passed forward. It is quite possible to put forward the alternative that the money supply should not be increased to accommodate the increase in prices, that aggregate demand should fall because of other exogenous factors, and that traders might be forced to absorb the vat, passing back its burden in lower factor rewards. As P.J.N. Sinclair, in his review of Value-Added Tax: Lessons from Europe, commented: “For example, suppose vat is replacing other, less efficient, indirect taxes, with an unchanged total yield. The reduction in the distortionary waste generated by the tax system behaves like a rise in aggregate output. The demand for money should therefore rise. So long as this is the only effect upon the demand for money, and the supply of money is unchanged, the overall price level should fall.”4

Whether or not forward shifting would be inflationary, as opposed to a once and for all change in the level of prices, depends not only on the possible offsetting changes in other taxes and an accommodating money supply but also on the reaction of wages, transfer payments, liquidity, and psychological effects.5

As Musgrave suggests,6 consider five policy settings:

  • Wages flexible, monetary policy permissive

  • Wages downward rigid, monetary policy permissive

  • Wages flexible, monetary policy stabilises product price level

  • Wages downward rigid, monetary policy stabilises product price level

  • Wages flexible, monetary policy stabilises factor price level.

He remarks that case (2) might be the most realistic and that, therefore, “an absolute price increase is the most likely result.”

For an increase in vat to be inflationary would not only require forward shifting and an accommodating credit policy but also a determination on the part of labor to maintain its real consumption, increasing wages sufficiently to offset the increase in prices due to VAT. The ability of labor to do this, of course, depends on numerous considerations outside the compass of a discussion of vat. One study shows that an increase in vat, even if fully offset by a reduction in income tax, so that a worker on the average wage has an unchanged real net income, can still produce a rise in all unionized wage rates.7 This increase in business costs could trigger the truly inflationary path of prices.

In addition to wages, there is the disturbing thesis put forward that recipients of transfer payments do not bear the burden of indirect tax increases, except perhaps in the short run, as governments will ensure that their real living standards are maintained.8 Thus, a significant burden may be transferred to income taxes, raising again, in an intensified form, the possible reaction of organized labor.

Though, as Alan Walters points out, referring to the 1979 U.K. increase in vat (from rates of 10 percent or 12 percent to a uniform 15 percent) fully offset by reductions in income tax, “it is difficult to see why such an obvious step—increase in certain prices, combined with downward pressure on the prices of excluded commodities, should lead to expectations of higher inflation (if, by inflation, we mean the persistent increase of prices year after year). Rationality would suggest expectations of lower price increases in future years.”9

The actual ways in which the authorities have tried to estimate the effect on prices range from almost back of the envelope guesses to sophisticated general equilibrium analyses. However, most have used partial equilibrium studies and some, depending on the taxes to be replaced and the data available, an input-output model for a static general equilibrium price determination. The first difficulty is to allocate the taxes replaced, and the vat (possibly at different rates), to the available input-output sectors. This inevitably means some substantial judgments, as there are likely to be fewer input-output sectors than taxed products. Having calculated the industrial-sectoral price changes these have to be reassigned to the constituent products that make up the household budget in the usual surveys and in the consumer price index; this demands another set of heroic assumptions. However, the anticipated price changes have matched the actual fairly closely and the apparent impact on inflation has been much as expected (see below).

Numerous countries, including Korea, Mexico, Morocco, New Zealand, Portugal, and the United Kingdom, have used partial equilibrium analysis to anticipate the effects of introducing vat and the results have, on the whole, turned out fairly well. Naturally, it is a bit difficult to know what influences what. Because the authorities have estimated the price changes that might occur, to some extent, those are the price changes that will be permitted to happen. As is shown below, governments undertake extensive actions to influence prices, from moral suasion to price controls, and these themselves reflect the judgments already made about expected price changes. Another difficulty, especially in complex substitutions, is that in some sectors, prices are calculated to fall (for example, the U.K. purchase tax and some Korean tax rates), and it is always easier for government to limit price increases than to require a price reduction. Usually in the estimates of the effect of the vat substitution on the consumer price index, it is better to err on the side of caution when the net effects of price increases are offset by price reductions. Experience shows price reductions rarely reflect the full tax reduction and some allowance should be made for this. In the Philippines, many traders did not roll back their prices by the full amount of the sales taxes replaced by vat. The vat base thus included the previous sales taxes. However, in some complicated cases, the outcome appeared close to the forecast. In Korea, processed food, beverages, furniture, and some medical goods were expected to increase in price, whereas confectionery, electricity and fuel, and clothing were forecast to fall. The net effect was designed to be approximately neutral. In the event, the authorities assigned little or no net price effect to the introduction of the vat. Indonesia expected a small (possibly 0.5 percent increase in the price level), and though there was a year on year consumer price increase of about 3.5 percent, only a trivial amount might be attributed to vat.

Most countries have not had the data to permit the use of general equilibrium analysis and, to date, none of the vats introduced have been tested against a general equilibrium prediction. The differences between this and partial analysis, for the vat, can seem large. Conventional incidence analysis involves the attribution of taxes replaced and the vat to commodities, but consumers will substitute out of taxed commodities, and general equilibrium could catch this adjustment.

The review that follows of the experiences of ten countries and their attempts to cope with the sensitive problems of price changes at the introduction of vat may strike U.S. readers as describing extraordinarily interventionist policies. It is difficult to imagine the United States introducing a vat in 1988 using a price freeze (Belgium, Korea, the Netherlands); price and wage controls (the United Kingdom); a national prices commission (Ireland); or profit margin control (Ireland, the Netherlands). The climate now is much more in favor of letting market forces operate and perhaps if these countries were to introduce vats now, such elaborate checks would no longer be used. Nevertheless, the more recent examples (for instance, Hungary, Portugal, and New Zealand) show that governments remain extremely concerned about the reaction of consumers as well as of voters to the effect on prices of introducing a vat or changing vat rates. Perhaps these brief case studies continue to have something to teach authorities about vat and prices. However, the basic message is that there is no reason to expect vat itself to be inflationary; its effects on prices, if any, will be once and for all.

Actual Policies10


Belgium changed over to vat on January 1, 1971. Previously it used a complicated form of cascade tax with a rate of 7 percent extending to all stages except the retail level. This was modified for certain industries by the adoption of a single rate, purporting to be equivalent to the aggregate of the amount payable at the various chargeable stages. Certain basic foodstuffs were exempt, and some other foodstuffs bore a single-stage tax at 7 percent. In addition, there was an invoice tax of 0.7 percent on turnover. Capital equipment was taxable.

Since the changeover to vat entailed a measure of credit for tax on capital equipment, and since a high proportion of Belgian industrial production was exported, it meant that for the same yield there would be a heavier tax burden on the domestic market. Moreover, basic foodstuffs, which were exempt before, became taxable under vat.

Four rates of vat were proposed, designed to correspond very broadly with the composite tax burden under the previous system. Basic foodstuffs were taxed at the bottom rate of 6 percent. A tax rate of 25 percent applied to goods chargeable at the former luxury tax rate, which had varied between 18 percent and 23 percent.

Price controls had operated in Belgium since 1945. A price freeze was in force for the first three months after the introduction of vat on January 1, 1971. Applications to change prices had to meet a three-month advance notice. This measure, and a firm policy on price rises afterward, limited increases to 2.3 percent in the first three months of 1971 and to 5 percent for the whole of that year, which was not much higher than for the preceding years.

An interesting aspect of the Belgian situation is that the standard rate originally proposed for vat was 20 percent with an intermediate rate of 15 percent. These rates were considered necessary if the degree of tax avoidance which allegedly existed under the earlier tax were to continue under vat. Business organizations strongly resisted those rates, and they were ultimately fixed at 18 percent and 14 percent, respectively. It was generally believed that during the period of uncertainty about the rates, many businesses widened their profit margins to protect themselves against any adverse effects that the higher rates proposed might have had on sales later on.


A comprehensive vat at 10 percent on goods and services was introduced in Denmark on July 3, 1967, to replace a wholesale tax at 12.5 percent, which contained several exemptions, including food, medicines, fuel, and services. Because of the extended scope of vat, its yield was estimated at double that of the earlier tax. It was accepted that the change was going to have a significant effect on prices. A reduction of income tax, especially for those at the lower levels of income, was promised and subsidy payments were applied to dairy products such as butter, cheese, milk, and cream. While the purpose of the subsidy payment was partly to relieve the consumer, the main object was to protect the producer in case home consumption of these items would fall if prices were to go up.

The effect of the vat on prices in Denmark can be calculated fairly accurately, because there were two price indices, one, the consumer price index, which took account of tax changes, and the other, the wage regulation index, which, since 1963, excluded taxes. Between April and October 1967, the consumer price index rose by 8 percent and the wage regulation index rose by 3.1 percent, so that the vat can be regarded as responsible for a rise of 4.9 percent in the consumer price index during the period. An increase of this order was expected because food and services were being taxed for the first time. No special price control measures were instigated. However, fairly large wage increases partially compensated for the erosion of real consumption.


From January 1, 1968, vat, which had existed in France for many years, was extended to the retail stage, and services were also brought within the system. Previously, vat was levied only to the wholesale stage, and there were separate taxes on retail sales and on services. These separate taxes were, of course, abolished. Bread, milk and other dairy products, and preserved meats were exempt under the earlier vat, and bread, milk, and certain foodstuffs in general consumption were exempt from the retail sales tax. All these items became chargeable under the new vat, so that a small increase in the consumer price index was expected.

In fact, during the first three months of 1968, the rise in the consumer price index was 1.15 percent, not all of which was attributable to vat—probably less than 1 percent. Analyses of the prices of 78 products were prepared, and these showed that the differences in prices before and after the changeover correctly reflected the tax changes.

Federal Republic of Germany

On January 1, 1968, Germany replaced its multistage sales tax on goods and services, which extended down to the retail stage, by a vat with a normal rate of 10 percent and a reduced rate of 5 percent on foodstuffs and agricultural products such as raw wool, raw hides, and timber. The change was to be revenue neutral and was not expected to have any significant effect on prices in general.

The rate of the replaced cascading tax was 4 percent (except at the wholesale stage where it was 1 percent, with an exemption for agricultural foodstuffs). The effective rate on a given item depended on the number of stages through which it passed before reaching the final consumer, but the broad implications of the change were that clothing and most foodstuffs would cost a little less, while services and building construction would cost more. Clothing would be taxed at 10 percent, compared with the earlier tax of 11 or 12 percent, and margarine, bread, potatoes, apples, beer, canned vegetables, sugar, eggs, coffee, and so on at 5 percent, compared with the earlier tax of about 6 or 7 percent. On the other hand, certain agricultural produce previously exempt, and meat formerly taxable at 4 percent, would bear vat at 5 percent. The largest tax increases were for services, of which gas and electricity were the most controversial.

Price changes directly attributable to vat worked out very much as expected, and accounted for a once and for all rise of 0.4 percent in the consumer price index. The largest increases were in the services sector, namely, hotels (4 percent), electricity (6.3 percent), gas (9.9 percent), public transport (5.2 percent), and cinemas (6.4 percent).

The vat was introduced at a time of mild recession in the German economy, and this was thought to contain undue price rises. The states (Länder) operated a price monitoring system and no further control on prices was deemed necessary. All the larger traders were well versed in the mechanics of vat, but there were problems with some small firms, which came under a special scheme where they were taxable at 4 percent without any credit for prior-stage tax. Some of these erroneously provided for tax at 10 percent in repricing on the changeover to vat. Consumer magazines and consumer organizations, however, kept a close watch on prices, and because of their vigilance unjustified increases were quickly corrected.

More recently, it has been held that the price level has risen by more than the tax increase in the wake of increases of the vat rate, which “suggests that enterprises use value-added tax increases as an opportunity to raise their own prices.”11


The changeover to vat in Ireland on November 1, 1972 was in itself price neutral because, except for a few minor items of little commercial significance, the rates chargeable under vat were calculated to produce precisely the same revenue as those under the replaced wholesale and retail sales tax on the estimated bases (even to the point of using some extraordinary rates to do so—such as 16.37 percent). Nevertheless, considerable public attention focused on the possible impact of the changeover on prices, and there was anxiety in case traders might try to widen margins.

The National Prices Commission, operating a system of price restraint, in its monthly report of March 1972 examined the likely outcome for prices of the changeover to vat; it concluded that any effect should not be significant. The Commission added that in all cases that would come before them after the introduction of vat, they would not recommend any price adjustment based on an increase in the absolute margin of profit or the application of that margin to a price that included any element of the tax paid at previous stages of production. If, following the changeover, prices were to rise appreciably it would be due to an unjustified reaction to vat by manufacturers and distributors or to cost increases other than vat. They announced that they would monitor the prices of a wide range of products during the period before and after the introduction of vat and would publish the results.

The autumn review of the Commission, published in January 1973, bore out the forecast that vat would have no significant impact on retail prices. While some price increases were reported, they were due either to seasonal factors or to passing on, with the approval of the Commission, higher costs of raw materials to the manufacturers. These were reflected in the rise in the consumer price index of 1.5 percent between August and November 1972.

Apart from the National Prices Commission, another organization, the Advisory Council on the Transition to vat, which was appointed by the Minister of Finance to consider any transitional problems arising from the implementation of vat, also directed its attention to prices. It carried out a publicity campaign through advertisements in newspapers, radio, and television, stressing that there would be no increase in the effective rate of tax on almost all goods and services, and recommending shoppers to report any cases of overcharging ascribed to vat to the Department of Industry and Commerce, which had increased its inspectorate to investigate any such complaints.

In addition, the daily newspapers opened “Prices Desks,” where consumers could report complaints or make inquiries about prices either in person or by telephoning. Where an instance of overcharging was brought to light, the name and address of the trader was published by the newspaper, as well as any observations he might offer to explain his case.

All in all, the Irish example might be considered a good example of a combination of government guidance on prices, official monitoring, and public awareness, successfully ensuring a smooth changeover to vat with little net effect on prices.


The Korean vat, introduced in July 1977, was the first vat in Asia and as such was a major innovation; after all, the Japanese had toyed with the idea of a vat since the 1953 Shoup mission, but had nervously postponed such a dramatic change in taxation. The Korean experiment was a brave initiative.

It was particularly courageous in that it involved the abolition of eight taxes, putting at risk a large proportion of revenue and, though the entire tax transformation was designed to be revenue neutral, there had to be numerous changes in relative prices. Naturally, there was the possibility, as mentioned above, that the changes might be asymmetrical so that prices would go up but other prices would not go down or would fall less than they ought. For instance, fuel, electricity, and clothing were expected to fall in price, while processed food, beverages, and furniture were expected to increase; the forecast change in the wholesale price index was 0.155 percent and the actual outturn was 0.061 percent; the retail price change was expected to be about nil and turned out to be an increase of about 1 percent. Overall, the limitation of the effect on prices, given the complex tax substitution and the great uncertainty, was an unqualified success.

The authorities consulted trade organizations before and during the design of the vat and looked for cooperation from trading organizations. Government control over factory and wholesale prices for some 250 products produced under monopoly and oligopoly circumstances meant that such prices could be held constant. Just before the introduction, a widespread campaign informed the public of recommended prices for many consumer goods. Finally, the vat was designed to be revenue neutral at a rate of 10 percent, but the authorities, nervous about the large amount of revenue at risk and the potentially disastrous consequences should there be a significant revenue shortfall, increased the statutory introduction rate to 13 percent. However, in what appears in retrospect to have been an inspired psychological move, just before July 1, 1977, the Government used a clause in the vat legislation allowing it to adjust the rate plus or minus 3 percent (supposed to be for purposes of macroeconomic stabilization) to reduce the initial rate to 10 percent. As it turned out, this 10 percent rate was noninflationary and produced somewhat more revenue than expected. Oddly enough, the statutory rate of vat remained 13 percent for many years, although the effective rate has continued to be 10 percent.


The vat was introduced in the Netherlands on January 1, 1969. Before that there was a cascade-type turnover tax which stopped at the wholesale stage. Under it, certain unprocessed foodstuffs and fuel for domestic consumption were exempt.

The vat was initially chargeable at a rate of 12 percent with a reduced rate of 4 percent applied, broadly speaking, to the areas that were formerly exempt. The changeover, therefore, entailed taxing basic foodstuffs and services for the first time. Although vat was expected to yield the same revenue as the system it replaced, an increase in the consumer price index was inevitable. The Netherlands Central Planning Office put the estimated rise at 1.3 percent. Agricultural products were expected to increase by 3.2 percent and other foodstuffs by 0.8 percent; textiles, clothes, and footwear by 5.1 percent; and furniture by 4.8 percent. On the other hand, decreases were anticipated in certain sectors, including a fall of 1.7 percent in transport and public utilities.

In the first three months following the changeover, prices rose by 5.2 percent, but the authorities considered that vat accounted for only 1.5 percent of this. The main factor in the rise appears to have been wage increases that were not matched by increased productivity. An important and trend-setting wage increase in the heavy metal industry provided for a rise of 6.5 percent; in addition, the higher social insurance contributions by employers brought the total increase in wage costs to 9.5 percent. There was an impression, too, that some manufacturers and traders were using the occasion of the changeover to vat to compensate themselves unduly for wage and other costs.

There had always been a system of voluntary price restraint in the Netherlands and no special measures were taken on the changeover to vat. However, to meet a later incipient inflationary situation, the Government brought in a price freeze on March 14, 1969. It compelled certain industries, whose price increases were considered excessive, to roll them back. This helped to stabilize the position. Moreover, no increased labor costs due to further pay awards could be passed on in prices and no increases in profit margins were permitted; but increases attributable to the earlier rises in wages were built into the new price structure. The extent of this increase, however, was not very considerable. In the first nine months of 1969, the consumer price index rose by 6.2 percent, and this was in part accounted for by foodstuffs, which became taxable for the first time under the new system (the average increase in price in the four preceding years was about 5 percent a year). Prices fell significantly after March 1969, and the freeze was relaxed gradually in June and September 1969.

New Zealand

The vat (known in New Zealand as the goods and services tax—gst12 was introduced on October 1, 1986. The authorities made many estimates of the likely impact of replacing the wholesale tax by a vat but, broadly, the once and for all expected increase in prices was about 5 percent.13

The authorities went to great lengths, using an extensive advertising campaign to reduce the public’s fear about price increases and to contain any possible attempt by traders to take advantage of the uncertainty to widen margins. It was estimated that about half of the 10 percent vat yield would be needed to replace the previous sales tax; this left about a further 5 percent net increase to be reflected in higher prices. As it turned out, some traders did increase their margins and the actual outturn for the first quarter, after the introduction of vat, was a price increase attributable to vat of 6.5 percent. The Government confidently expected the rate of inflation to return to the previous trend in successive quarters, as it did, so that the introduction of vat can be seen, not as inflationary, but as a once and for all blip.


Sweden brought in vat on January 1, 1969, at a rate of 10 percent on tax-inclusive prices. It replaced a retail sales tax with the same rate. As, however, capital equipment was taxed under the retail tax, but tax free under vat, the yield from vat was expected to be less than from the former tax. The loss was made good by a 1 percent payroll tax.

The changeover was effected smoothly, and was looked on as being little more than a change in the method of collecting sales tax. The information for traders was entrusted to a large degree to the trade organizations, which trained their members on the operation of the new system.

The impact on prices was almost neutral, because the rate of vat was identical with that of the previous retail sales tax. During the first eight months of 1969 the consumer price index rose by 2.3 percent, compared with an average annual increase of 5 percent in each of the preceding four years.

No special price restraint was instigated at the time of the changeover, but when the rate of vat was raised to 15 percent (again on tax-inclusive prices) in January 1971, control measures were introduced, limiting price rises to the amounts justified by the change in rate.

A later econometric study suggests that the vat is reflected in higher prices within a quarter of a year. It is held that the 1971 tax increase (from 15 percent to 20 percent) increased the consumption deflator by 2.5 percent, compared to no tax increase; “however, the model does not include the institutional frame of labor market negotiations which is very important for Sweden.”14

United Kingdom

The vat was introduced in the United Kingdom on April 1, 1973 to replace two taxes: (1) the purchase tax, a single-stage selective sales tax then chargeable at three different rates (25 percent, 18 percent, and 11.25 percent, and these rates represented substantial reductions from previous levels in preparation for the introduction of vat) on the wholesale values of many goods; and (2) the selective employment tax, which was assessed not on the price of goods but on the number of employees and the nature of the business.

Charges that vat would be inflationary and regressive were made from the time the new tax was announced. Fear was expressed, in particular, that unless full relief were given for inventory stock held at the changeover, retailers would have to raise their prices unduly to recoup themselves for the double taxation burden. Anxiety on this score was removed when the Government eventually agreed to give full relief.

In May 1972, the National Institute of Economic and Social Research forecast that a vat, which was to raise exactly the same amount of revenue as the taxes it would replace, would lead to an increase of 0.5 percent in the consumer price index, and that a 10 percent vat with the coverage then envisaged might increase consumer prices by more than 1 percent.

A price and wage standstill instituted in November 1972 was to be succeeded by the “stage 2” relaxation at the time of the introduction of vat in April 1973. Toward the end of 1972, consultations on price control began between the government departments directly concerned with price restraint. The outcome was the Counter-Inflation Bill, introduced in Parliament early in 1973. It contained a clause about retail prices in the period of transition to vat, drafted after consultation with the Confederation of British Industry and the Retail Consortium, on the basis that a special temporary price control could be set in motion by an individual consumer’s complaint about a particular price. The complaint would lead to an investigation of the facts; if it were found that the price change did not correctly reflect the tax changes, a notice would be served on the shopkeeper requiring him to correct the price. Disregarding the notice would be an offense. This approach was adopted because public anxiety about the impact of vat on prices had to be allayed. It was recognized that, initially, many retailers might be genuinely perplexed and that not all instances of apparent overcharging could be regarded as profiteering. Some discretion and flexibility was considered necessary.

The U.K. changeover to vat is unusual in that few countries have made the transition from taxing goods at high rates to taxing them at a uniform rate. (In some cases, the taxes were extremely high; jewelry, a few years before vat, could have borne tax at 60 percent of its wholesale value.) The changes in relative prices were substantial, even though the size of the rate differentials had been narrowed in anticipation of vat. The January before vat was introduced provided a good illustration of the shoppers’ dilemma. Should they buy goods at winter “sale” prices or wait till April when prices ought to be reduced under vat?

The basic scheme to provide the public with facts on which to make judgments was to provide lists of goods and services with the price changes that could reasonably be expected for each, quoted first in ranges of pence per pound; then a typical price before vat, and the typical change (or range of typical changes) in that price. The first government advertisements, full newspaper page length, appeared in mid-March with the lists divided between three panels headed “about the same,” “up,” and “down.” Calculating the price changes, item by item, and checking whether the typical prices really were typical entailed inspection of shop shelves, catalogs, and advertisements, a task undertaken by the tax authorities. The lists were shown in advance to the Confederation of Industry and the Retail Consortium to give them an opportunity of commenting on the typical prices, the assumptions made about retail markups, and the detailed effects of abolishing the selective employment tax.

Apart from the press, radio, and television publicity, a leaflet giving the same information about prices was prepared with the title “Your Guide to vat Prices,” and 5.5 million copies were distributed free through post offices.

The press reports of the time show that, thanks to the effectiveness of the government publicity campaign, the changeover went smoothly, with, as anticipated, a large number of consumer durables becoming cheaper, and many services rising by up to 10 percent. There were complaints that some firms were not passing on the benefit of the abolition of the selective employment tax, but immediate attention was given to this, and the authorities said that over the country as a whole, the movement of prices and changes were very close to the estimates made when the publicity campaign was launched a month previously.

The Retail Consortium, representing about 90 percent of large retailers, had few reports of difficulties, and this is again ascribed to the thoroughness of the government advertising campaign.

The Central Statistics Office with the Treasury and Customs and Excise undertook a study of the increase in the retail price index over the four-week period from March 20. The April retail price index showed an increase of 1.9 percent on the previous month. Of this, it was estimated that higher food costs accounted for about 0.75 percent and raw materials another 0.5 percent. Since neither of these were affected by vat, the result of the transition to vat was less than 0.7 percent. Ministerial statements at the time said that the effect of the tax changes on prices had not been significant.

A survey carried out by the Consumers Association, whose representatives visited 800 retail shops three weeks before the introduction of vat and again in the second week of April, said that of 2,188 prices checked, 40 percent went up less than expected, while 20 percent rose more than expected or were reduced by less than the amount set out in the shoppers’ guide. They found there was more evidence of “fiddling” in the services, such as restaurants and cinemas, than in the shops.

Reports to the responsible ministries showed that during April the government inspections of complaints about prices proved that about 70 percent of them were unjustified and, in another 25 percent, voluntary reductions were secured as a result of the inspectors’ inquiries. In only 12 cases, where the complaints were considered justified, was the trader unwilling to alter his price. In these cases, the inspectors issued statutory notices restricting the prices in accordance with the provisions of the Counter-Inflation Act, 1973.

The success of the U.K. introduction of vat, despite a most difficult set of relative price changes, is principally due to the care the authorities took in cooperating with industry and the public to anticipate and meet their major concerns about vat. The Government published periodic vat bulletins for traders and tried to pay an educational visit to each before vat was introduced. The monitoring of price changes by official agencies, consumer watchdogs, and newspapers undoubtedly provided a coverage that convinced everyone that the Government was serious about controlling unjustified price increases and ensuring price reductions when traders had been relieved of tax. Table 10-1 summarizes the main effects on prices of the introduction of vat.

Other Empirical Evidence

Changes in the consumer price indices, credit, and wages have been examined in 35 countries for two years on each side of the date vat was introduced; another 6 examples of rate changes in existing vat systems were investigated.15 The actual outturns were compared to four hypotheses, first, that the vat could lead to a once and for all shift in the trend of the consumer price index; second, it might generate an accelerated increase in the rate of growth of the consumer price index (that is, inflation); third, there could be a shift and acceleration in the consumer price index; and finally, there might be little or no effect. Table 10-2 summarizes the results.

Table 10-2.Countries Allocated to Categories for Effect on Consumer Price Index
Introduction of VATChanges in VAT Rates
CasesOn data

Considering all

On data

Considering all

ShiftBoliviaDenmarkDenmark, 1977Denmark, 1977
AccelerationArgentinaGuatemalaNetherlands, 1971
United Kingdom
Shift plus accelerationHonduras NorwayNorwayIreland, 1980Ireland, 1980
Little or no effectAustria BelgiumArgentinaBelgium, 1983Belgium, 1983
BrazilBelgiumNetherlands, 1971
ColombiaBoliviaNetherlands,Netherlands, 1973
Costa RicaBrazil1973Netherlands, 1976
Côte d’IvoireChileNetherlands,
IndonesiaCosta Rica
KoreaCôte d’Ivoire
United Kingdom
Source: Alan A. Tait, “The Value-Added Tax: Revenue, Inflation, and the Foreign Trade Balance,” paper presented at World Bank Conference on Value-Added Taxation in Developing Countries, Washington, April 2l–23, 1986.
Source: Alan A. Tait, “The Value-Added Tax: Revenue, Inflation, and the Foreign Trade Balance,” paper presented at World Bank Conference on Value-Added Taxation in Developing Countries, Washington, April 2l–23, 1986.

An interesting outcome was the extremely clear indication of a pronounced shift in eight of the countries examined. However, only in one case (Norway) was the shift unambiguously combined with an acceleration in the rate of change of the consumer price index. In another (Bolivia), after looking at other evidence, it was concluded that the shift was not due to the vat. So in all, there were 7 cases where the vat created a shift in the trend of the consumer price index. In 5 countries, there was an acceleration; in 1, a shift and an acceleration; and in the rest, 22, or 63 percent of the countries, there was little or no effect of the introduction of vat on the consumer price index.

Apart from the introduction of vat, evidence of the effect of the tax on prices might be judged from changes in the rates of existing vats. Examination of six rate changes (see Table 10-2) does not suggest any automatic link between vat rate changes and inflation. Only one country (Ireland) suffered accelerated price changes, one country (Denmark) exhibited a shift, and the other four showed little or no effect.16

While some commentators have associated vat rate changes with inflation,17 the weight of evidence seems to be against it. Indeed, practical common sense suggests the opposite: “Price increases remove inflationary pressure; they do not add to it. Some light may be shed on the issue by pursuing the symmetrical argument for lower indirect taxes. If it were possible to excite inflationary expectations by a switch to indirect taxes, would it not follow that one could subdue such expectations by switches from indirect taxes to income tax? Or perhaps to more borrowing to finance some subsidies?”18 In any case, further assumptions have to be made on an accommodating monetary policy and on the adjustments in wages and transfers.

For an examination of these issues and a different presentation of evidence, see Tait (1986).

“Portugal: Effects of VAT,” World Tax Report (1986, p. 9).

Ture (1972, p. 100).

Sinclair (1983, p. 382).

See the discussion in Tanzi (1983).

Musgrave (1986, p. 215).

Sampson (1986, pp. 87–91).

Browning (1978) and Browning and Johnson (1979); also see United States, Department of the Treasury (1984, Vol. 3, pp. 91–92).

Walters (1986, p. 77).

This section has benefited greatly from help given by Seamus Duignan.

Dengel (1987, p. 279).

Suspected named after the Assistant Secretary to the Treasury, Graham Scott, as the “Graham Scott Tax”—see Douglas (1987, p. 209).

See, for instance, Douglas, Statement on Taxation and Benefit Reform 1985 (1985, p. 10); see also Stephens (1987, p. 336).

Andersson (1987, p. 49).

Tait (1981, pp. 38–42) and a more extensive and up-to-date discussion in Tait (1986).

This is borne out by Sijbren Cnossen’s reference; see Cnossen (1981, Table 5 and pp. 244–46).

Hemming and Kay (1981, p. 82).

Walters (1986, p. 77).

    Other Resources Citing This Publication