Printer Friendly

The value added tax: its impact on saving.

*Ken Militzer is Chief Economist of AT&T and Ilona Ontscherenki is District Manager-Financial and Policy Analysis at AT&-T, Morristown, NJ. The authors are indebted to Gerald DuGuay for research support and helpful comments. [1]See footnotes at end of text.

A value added tax (VAT) has been touted as a way to stimulate U. S. saving. However, a cross-country analysis of OECD data for twenty-three countries shows that saving rates are not higher in countries that rely more heavily on a VAT for revenue. Therefore, enacting a VAT in the U. S. is not likely to increase its saving rate. Furthermore, other features of a VAT suggest that it would not be a good tax for the U. S.

ONE ADVANTAGE that many supporters cite for substituting a VAT for income taxes is that a VAT - as a tax on consumption - would stimulate the saving rate in the U. S., which is among the lowest in the world.[1] They argue that the income tax discriminates against saving by taxing consumption once, but saving more than once, because the income from saving is also taxed. Supporters of a VAT say that taxing consumption instead of income would reduce or eliminate that discrimination and, therefore, stimulate saving.[2]

Although it seems theoretically sound to contend that placing greater reliance on a VAT or consumption taxes to raise government revenue would result in increased saving, we have been unable to find any empirical evidence that that actually happens in practice. For example, if personal saving were responsive to changes in income tax rates,[3] the U. S. saving rate should have increased following the supply-side tax cuts during the first Reagan Administration. Those cuts totaled 25 percent and included a reduction in the top tax rate on income from capital (so-called "unearned income") from 70 percent to 50 percent. However, the personal saving rate in the U.S. declined in the years following those tax reductions, rather than increasing.

Moreover, data for twenty-three OECD countries[4] indicate that neither consumption taxes in general, nor a VAT or retail sales tax in particular, have any impact on the rate of saving. This conclusion is evident from the scatter diagrams shown on Charts 1 and 2.

Charts 1 and 2 compare, for the year 1986, the rate of national saving (as a percent of GDP) for each of twenty-three OECD countries with each country's relative use of consumption taxes[5] (Chart 1) and VAT and retail sales[6] taxes (Chart 2), measured as a percent of total taxes in each country. As can be readily seen, there is no relationship between the relative use of consumption taxes, or of VAT and retail sales taxes, and the rate of saving in those countries. That visual evidence is reinforced by more formal statistical analysis. Regressions calculated for the relative use of consumption taxes and the rate of saving, and the relative use of VAT and retail sales taxes and the rate of saving, for the twenty-three OECD countries in 1986 showed no statistically significant relationship between the national saving rate and either of the tax series.[7]

Of course, the rate of national saving includes the impact of government surpluses or deficits on the overall saving rate. To see if a VAT might affect private saving differently from national saving, we also compared the relative use of consumption taxes, and VAT and retail sales taxes, with the rate of private saving. Unfortunately, the OECD publishes data for private saving for only seventeen of the OECD countries.[8] For those seventeen countries, though, there again was neither a visual nor a statistical relationship between the use of those kinds of taxes and the rate of private saving in 1986.

Because a test of just one year might not be conclusive, we ran the same regressions for data for 1965, 1970, 1975, 1980, and 1985. In no case, however, did we find a statistically significant positive relationship between consumption taxes and saving rates, or between VAT and retail sales taxes and saving rates.

If consumption taxes, or VAT and retail sales taxes, really were effective in promoting saving, some relationship between the relative use of such taxes and the rate of saving should be discernible among these twenty-three OECD countries. The complete lack of any such relationship - including its total lack among the three major countries highlighted in the charts (Japan, with no VAT or retail sales taxes in 1986 had the highest saving rate of any OECD country except Luxembourg) - indicates that a VAT is not an effective way to stimulate saving.

OECD data also indicate that shifting income taxes to consumption taxes would not promote saving. If income and profit taxes effectively discouraged saving, the rate of saving should be lower in countries with high income and profit taxes and vice versa. To test whether that might be the case in practice, we ran regressions for the relative level of income taxes and the national saving rate (twenty-three countries) and the private saving rate (seventeen countries) for the years 1965, 1970, 1975, 1980, 1985, and 1986. In no case was there a statistically significant relationship between the use of income taxes to raise revenue and the level of the saving rate in those countries. That lack of any relationship is shown on Chart 3 for private saving for the year 1986. With no evidence that greater relative reliance on either consumption taxes or income taxes has affected the rate of saving in these countries, there is no reason to believe that shifting the tax burden from income to consumption taxes in the U. S. would have any impact on the rate of saving in this country.


As a further test of the impact of consumption and value added taxes on saving, we performed a number of tests to determine whether changes in the relative use of consumption taxes, of VAT and retail sales taxes, and of income taxes, in the OECD countries could be associated with changes in their rates of saving. We did this by comparing changes in those taxes as a percent of total taxes with changes in both the rate of national saving twenty-three countries) and the rate of private saving. For national saving, these tests were made for the twenty years from 1965 to 1985 and for the fifteen years from 1970 to 1985. For private saving, they were made for the fifteen years from 1970 to 1985, because the data needed to derive private saving are not available for 1965.

In no case could we find any regressions in which there was a significant positive statistical correlation between changes in the rate of national or private saving and changes in the relative use of consumption taxes or of VAT and retail sales taxes. Similarly, we found no relationship between changes in the use of income taxes and changes in the rates of saving in the OECD countries.


National and private saving were used in the analyses discussed above, because they are the broadest measures of a nation's saving rate. However, it might be argued that VAT, consumption, and income taxes would have a more direct effect on the rate of household (i.e., personal) saving. Therefore, we ran all of the same regressions discussed above for household saving.[9] Again, we found no statistically significant relationship between the relative use of those kinds of taxes and the rate of household saving in any of the years studied, nor did we find such a relationship between changes in the use of such taxes and changes in the rate of household saving.

In sum, all of these cross-sectional analyses of the OECD countries - the static analyses of the years 1965, 1970, 1975, 1980, 1985 and 1986, and the dynamic analyses of the 1965-85 and 1970-85 time periods - point to the same conclusion. A nation's saving rate is not influenced by the degree to which consumption taxes or income taxes are used to finance government spending. Because increasing private saving is one of the main advantages asserted for a VAT, we see no reason that a VAT should be adopted by the U.S.


Some advocates of a VAT say that such a tax would help increase U. S. exports, because it would be rebated on goods that are exported, and that it would help reduce U.S. imports, because it would be added to the price of imports. In contrast to that treatment, corporate income taxes are not permitted, under the rules of international trade, to be deducted from exports, making the use of a VAT favorable for a country's international trade.

That analysis, however, applies only if the VAT were substituted for the corporate income tax. Because the corporate income tax accounted for only 10.4 percent of federal revenues in 1988, any significant use of a VAT in the federal tax system would undoubtedly go well beyond such a substitution. But if a VAT were either an additional tax or a substitute for personal income taxes, it would not help our international trade:

1. As an additional tax, it would not improve U. S. competitiveness. Rebating a VAT on exports would leave prices of exports as they were before, whereas adding it to imports would only cause prices for imports to reflect the same VAT carried by domestic products.

2 .As a substitute for personal income taxes (which accounted for 44 percent of federal revenues in 1988), a VAT would have no impact on our international competitiveness, because personal income taxes do not directly affect either producers' costs or their prices.

Moreover, some analysts believe that even the hypothetical advantage of substituting a VAT for the corporate income tax may be spurious. International exchange rates keep international trade and capital flows for each nation in balance by adjusting for all sorts of differences among nations. If one country's trade situation were improved by the substitution of a VAT for the corporate income tax, that advantage should be offset, at least over time, by a countervailing move in the value of its currency. Thus, the OECD has said that a consensus seems to have emerged that the impact of consumption taxes on a country's international competitiveness "in these days of flexible exchange rates . . . is not an important issue."[10]

Thus, two major advantages that supporters of a VAT claim for it do not appear to exist in practice. The empirical evidence indicates that a VAT does not promote saving, while the prospect that a VAT would help our international trade both is questionable and, in any case, applies only under very narrow conditions. At the same time, a VAT has many disadvantages, discussed below, that make it a very poor tax.


Even supporters of a VAT agree that a VAT has a number of disadvantages. It is inflationary, regressive, complex, costly to administer and comply with, and, in practice, discriminatory. Moreover, it would place the federal government in competition with state and local governments for a kind of tax that, in the U. S., has previously been the exclusive domain of state and local governments.


As a tax on consumption, a VAT would be immediately inflationary, i.e., it would raise consumer prices by about the same amount as the tax rate that was implemented. Although that would be a one-time event, there would probably be secondary effects, due to various COLA clauses and to efforts by workers to bargain with their employers to make themselves whole for the inflation. Moreover, even as a one-time event, the inflationary impact of a VAT would have an adverse impact upon the real value of existing savings.

Although supporters of a VAT sometimes argue that it has not been inflationary in Europe, that is because most European countries enacted VATs as a replacement for transaction taxes, which had a similar impact on prices. Where that was not true, the introduction of a VAT did prove to be inflationary. For example, in Denmark the price level increased by 5 percent because the VAT was larger than the tax system it replaced.[11] Japan's inflation rate also surged recently with the implementation of a 3 percent VAT.

Supporters of a VAT have also argued that its inflationary impact could be mitigated or even avoided by the proper use of monetary policy. Unfortunately, monetary policy combats inflation by reducing the rate of economic growth. Thus, the nation would be faced with the choice of having monetary policy accommodate the inflationary impact of the VAT by allowing money supply to increase, or having monetary policy fight the inflationary impact of the VAT at the cost of slower economic growth or even a recession. We do not believe that the latter choice would be a satisfactory alternative, so we do not believe that the inflation resulting from a VAT would be avoided.


Because low-income persons consume a higher proportion of their income than high-income persons, a VAT would be regressive, i.e., the VAT would take a higher percentage of the income of low-income persons than it would of high-income persons. Although everyone agrees that regressivity is undesirable, supporters of a VAT argue that the problem of regressivity could be handled by an income-tax credit for low-income persons. However, such a credit certainly would not provide a perfect offset to regressivity, and it would also add complexity to the income-tax code. Moreover, in practice, countries with a VAT usually have attempted to offset its regressivity by exempting certain products and/or by using variable rates for different products. As discussed below, however, that technique both adds to the complexity of a VAT and makes it discriminatory among industries.


Partly because of efforts to offset its regressivity, in practice VATs have proven to be very complex. The use of exemptions and of different rates for different products have created tremendous definitional problems in many countries abroad. That has both complicated the political process and added to the cost of administering a VAT by both businesses and government.


An ideal VAT - one with no exemptions and one rate - would have the advantage of being a neutral tax among different industries. That reason is frequently cited as an advantage for the VAT over an income tax, which some observers say discriminates against capital-intensive industries.

In practice, however, VATs do not have single rates and they do not apply to all products. Rather, they tend to have a number of rates and many exemptions. In Europe, for example, they tend to have three rates: standard rates, which range from 12 percent in Luxembourg, Spain, and Turkey, to 25 percent in Ireland; lower rates O to 17 percent), which are used to combat regressivity and usually cover items such as food, medical care, books, electricity, etc.; and higher rates (usually over 30 percent), which typically apply to luxury items, such as automobiles, televisions, VCRs, jewelry, and furs.[12]

As a result of this multiplicity of rates, in practice VATs are probably even more discriminatory among different industries than the corporate income tax.

Administrative Costs

Because a VAT would be an entirely new tax within the U.S., it would require additional costs to administer both for businesses and the government. In Europe, those costs range from 0.32 percent of revenue in Norway to 1.09 percent in Belgium.[13] However, the standard rate for European VATs ranges from 12 percent to 25 percent. Administrative costs as a percent of revenue would be much higher for the lower rates of VAT that are being talked about for the U.S., and, of course, industry costs would be in addition to those government costs.

Although supporters of a VAT argue that each of those problems could be addressed in one way or another, provisions that would solve one problem frequently would aggravate another. For example, reducing regressivity frequently would add complexity to the tax code, increase administrative costs, and create discrimination among industries. Combating the inflationary impact through monetary policy would reduce the rate of economic growth.


Because it is so broadly based, the VAT can raise very large amounts of revenue. Moreover, if no exemptions are allowed, it can do so at relatively low tax rates. That has caused a VAT to be praised by some as an easy way to reduce or eliminate the federal deficit and feared or condemned by others as a potential money machine that would lead to greatly increased federal spending.

To test whether a VAT has the characteristics of a money machine in practice, we again looked at the experience of the OECD nations. We looked at the relationship between VAT and retail sales taxes as a percent of GDP and total government outlays as a percent of GDP, to see if government spending is higher in countries that more extensively use VAT and retail sales taxes than it is in countries that do not. The scatter diagram showing that relationship in 1986 for the seventeen OECD nations for which data on government outlays are available [14] is contained in Chart 4. As can be seen, there is a significant relationship R2=0.55), with government spending as a percent of GDP (vertical scale) increasing as the use of VAT and retail sales taxes (horizontal scale) increases. A similar although weaker relationship was found between total consumption taxes and government spending (R2 = 0.31), while virtually no relationship was found between the level of government spending and the use of income and profit taxes (R2 = 0. 15).

This analysis of OECD data showing that those countries that rely most heavily on a VAT or other consumption taxes tend to have larger government sectors than those that rely less heavily on them is consistent with fears that a VAT is a money machine whose use could lead to increased government spending.[15]

A characteristic of a VAT that adds to fears that it would be a money machine is that it can be a hidden tax. Although the VAT needs to be shown separately on invoices for business sales (to facilitate the claiming of credit by businesses for the VAT paid to their suppliers), in some countries the cumulative VAT paid by consumers is not shown separately. Therefore, some observers fear that the tax rate on a VAT could be raised more easily than other tax rates.


Although a VAT is touted as a tax on consumption, in practice it is imposed directly on businesses. It becomes a tax on consumption to the extent that businesses are able to shift the tax to consumers in the final price of their products, similar to the way many analysts believe that the corporate income tax is shifted to consumers in the final price of the products of corporations paying it. To the extent that businesses are unable to shift a VAT fully to consumers by including its entire amount in the final price to consumers, the VAT is then borne by businesses.

Whether businesses would have a higher tax burden under a VAT than they do under the corporate income tax is not entirely clear. Some analysts believe that the corporate income tax falls entirely on capital (i.e., on business rather than on consumers). However, many observers believe that much of it is shifted forward to consumers in the price of the final product, especially in the long run. Such shifting presumably is not complete, however, because not all corporations pay the corporate income tax (some companies lose money, at least in some years) and companies that do pay corporate income taxes do so at different effective rates. Therefore, it seems likely that not all of the corporate income tax is shifted to consumers.

Most analysts presume that a VAT would be fully shifted to consumers, because all businesses producing the same product would pay the same rate. However, because of its potentially huge size relative to the corporate income tax, a VAT would have a very adverse impact on business unless it were completely shifted.

How much of a VAT would be shifted to consumers and how much of the corporate income tax is shifted to them is not something that can be determined scientifically. However, there is little question that all of the personal income tax is borne by consumers. Therefore, it seems to us that the ultimate cost of businesses of raising revenue by means of a value added tax would not necessarily be any less than raising an equal amount of revenue through the combined personal and corporate income tax.


All of the empirical evidence that we can find demonstrates that a VAT does not accomplish the major objective that supporters claim for it, i.e., to increase saving. Moreover, it may also not improve U.S. competitiveness. However, a VAT does have many serious drawbacks that could not be easily overcome. Therefore, we believe it would be a serious mistake for the U.S. to enact a VAT.


1. The Saving Solution," by Thomas E. Swanstrom in Business Economics, july 1989, describes recent U. S. saving trends.

2. In putting forth that argument, supporters of a VAT seem to ignore the difference between taxing saving itself and taxing the income from savings. They also seem to ignore that there are also additional taxes on consumption, and they ignore the discrimination of a different sort that would result from taxing income from labor, but not income from capital.

3. Interestingly, when changes in income taxes have been used contracyclically in the U. S., the changes have been made to stimulate or restrain spending, not saving.

4. These include Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Japan, Luxembourg, Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, Switzerland, Turkey, United Kingdom, United States. Data sources: OECD, Revenue Statistics, Economic Outlook.

5. The OECD definition of consumption taxes includes retail sales, excise, and certain other similar taxes, so that even countries like the U.S. and Japan, which did not have a VAT in 1985 or earlier, still collected sizeable amounts of consumption taxes.

6. In terms of their impact on the rate of saving, there should be no difference between a VAT and a retail sales tax, so they are combined in this analysis.

7. In these and subsequent charts, the three major industrial countries - the U.S., Japan, and Germany are highlighted to see if they might show a noticeable relationship between the use of these taxes and their rates of saving that is not shown by the other OECD countries. As can be seen, they do not.

8. The OECD does not publish data for private saving for ireland, Luxembourg, New Zealand, Portugal, Switzerland, and Turkey.

9. The OECD does not publish household saving data for Denmark, Ireland, Luxembourg, New Zealand, Portugal, and Turkey. Data for 1970 were unavailable for Australia and Norway.

10. OECD, Taxing Consumption, 1988, p. 32.

11. Tait, Alan A., Value Added Tax: International Practice and Problems, International Monetary Fund, 1988, p. 201.

12. OECD, Taxing Consumption.

13. OECD, Taxing Consumption, p. 204.

14. The OECD does not publish such data for Ireland, Luxembourg, New Zealand, Portugal, Switzerland, and Turkey.

15. Although Charles E. McLure, Jr., a VAT proponent, has indicated that the money machine argument is valid ("The VAT and Other Revenue Sources," Business Economics, October 1988), not all analysts agree. J. Q. Stockfisch, in a chapter contributed to the book, The Consumption Tax, edited by Charls E. Walker and Mark A. Bloomfield, also noted the tendency for countries with a VAT to be high-spending countries. Comparing the experience from 1971-73 to 1981-83 of twelve OECD countries that had a VAT with twelve OECD countries that did not, Stockfisch found that the VAT countries, on average, had larger government expenditures relative to GDP to start with (36.0 percent vs. 24.0 percent), had larger percentage point increases in government expenditures during that period (13.9 percent vs. 8.8 percent), had a larger adverse swing in government saving or deficits as a percentage of GDP -6.1 percent vs. - 3.7 percent), and had larger increases in taxes as a percentage of GDP (7.8 percent to 5.1 percent). However, Stockfisch does not believe that that experience supports the contention that a VAT is a money machine. Rather, he concludes that, ". . . VAT countries appear to have been big government spenders before the advent of VAT, and they have continued to behave that way . .. " He notes that consumption taxes as a percent of total taxes declined over that period, on average, for both VAT and non-VAT countries.
COPYRIGHT 1990 The National Association for Business Economists
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1990 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:comparison of 23 countries who employ value added tax
Author:Militzer, Ken; Ontscherenki, Ilona
Publication:Business Economics
Date:Apr 1, 1990
Previous Article:Public policy towards corporate restructuring.
Next Article:Policy response to countertrade and the U.S. trade deficit: an appraisal.

Terms of use | Copyright © 2016 Farlex, Inc. | Feedback | For webmasters