The Fraser Institute: Tax Reform in Canada:  Our Path to Greater Prosperity

A Fraser Institute Conference,
October 11, 2001, Toronto, Ontario, Canada

Why there should be no Capital Gains Tax

Herbert Grubel
Professor of Economics (Emeritus), Simon Fraser University
Senior Fellow and David Somerville Chair, The Fraser Institute
herbert.grubel@home.com

A paper to be presented at the conference "TAX REFORM IN CANADA: Our Path to Greater Prosperity", sponsored by the Fraser Institute, held in Toronto on October 11, 2001.

First draft – September 21, 2001

Canada's taxation policies should be under constant review in the light of new empirical information about its performance and the merit of alternatives.  One would expect that after a long period when few changes were made, some major reforms would be very beneficial.  In addition, fiscal surpluses, as had been predicted for some years into the future, tend to offer the opportunity to reform the system without the constraint of neutral revenue effects, which would make it possible to have every conceivable interest group emerge with lower tax burdens.

Alas, the economic slowdown of the last year and the terror attacks on September 11 and their aftermath have eliminated the prospect of fiscal surpluses.  However, prosperity and surpluses are certain to return at some point in the future.  It is therefor important to consider now what changes in tax policy are warranted at this time.  Some of these changes might even be justified in a fiscal environment without surpluses.

I am also hopeful that the ideas for tax policy changes presented at this gathering will stimulate the government into having a major review of its own.  Such a review is long overdue and it would, at relatively little cost, force a broader public debate on the issues than we can generate with the much more limited resources at our command.

Nearly 30 years after the imposition of the capital gains tax in Canada in 1972, the time has come to reassess the empirical judgements, which provided the rationale for the tax.  There are four such empirical issues.  First, how large is the loss in income created by the tax; second, how much does the tax equalize after-tax incomes; third, what are the revenue implications of eliminating the tax and fourth, how difficult is it to prevent shifting of taxable income into non-taxable capital gains.

I.       Efficiency Implications

There is widespread agreement among experts that the capital gains tax, along with other taxes on capital, reduces the size of Canada's capital stock and therefore labor productivity and overall living standards.  The reasons for this result are readily summarized in a simple analogy.  Taxes on consumption and income are like taxes on the fruits of a tree.  Sharing the fruits with the government does not diminish future harvests.  Taxes on capital, on the other hand, damage the tree and reduce the size of harvests in the future.  There are two basic ways in which the capital gains tax causes this damage.

Lower Capital Stock

First, in an integrated global capital market, investors equalize in all countries the risk-adjusted rates of return after taxes.  If a small country, like Canada, imposes a capital gains tax, the actions of investors raise the pre-tax return on capital until the after tax return is made once more equal to that in the rest of the world.  The actors in this process are not just foreigners but also Canadians, who have the option of placing their money anywhere in the world.

          The higher the interest payable on capital, the less industry can afford to borrow.  As a result, Canada's total capital stock is smaller than it would be without the capital gains tax.  Since productivity of labor is an increasing function of the capital stock, the smaller amount of capital in Canada is associated with lower output and living standards.

The imposition of the capital gains tax tends to have its impact only slowly since it takes time for the capital stock to fall.  In addition, the effect tends to be hidden because the growing overall wealth in the world continuously raises the stock of capital and it is difficult to notice that this growth would have been greater in the absence of a capital gains tax.

The Lock-in Effect

Second, the capital gains tax causes damage to the productivity of capital by creating a so-called lock-in effect, which arises because the tax creates a wedge between the private and social rate of return of capital.  In the absence of the tax, an investment in a company with expected low earnings causes it to be sold and the funds to be reinvested in other companies yielding a higher rate of return.  This turnover of capital is part of the dynamism, the process of creative destruction, of an efficient economy.  In the absence of a capital gains tax, small differences in return causes such reallocations of capital as long as transactions costs are covered by the higher yield.  However, if the sale of an asset leads to a capital gain, this efficient process of reallocating capital is short-circuited.

          Consider an investor who holds an asset worth $100.  The future rate of return is assessed to be 9 percent and thus brings $9 annually.  In the absence of the capital gains tax, the asset would be sold and reinvested in a comparable company assumed to exist and with a return of 10 percent, bringing an annual income of $10.  But now consider that the sale of the asset gives rise to a $20 capital gains tax obligation, leaving only $80 for reinvestment in the higher yielding asset.  If the investor holds on to the low yielding asset, the income is $9, more than the $8 earned by selling it, paying the capital gains tax and reinvesting in the higher yielding asset.  The rational owner of the investment subject to the capital gains tax keeps his funds where they are and is privately better off than if the asset were sold.  However, society is worse off because funds are prevented from flowing from the lower into the higher yielding investment - $1 per $100 in the preceding example.

          The quantitative effect of the lock-in effect is difficult to estimate since the government has roll-over provisions for some types of investment under which the capital gains tax is not payable if the reinvestment takes place under a strict set of rules.  Assets held in tax-sheltered investments like RRSPs do not pay taxes on capital gains.  And, of course, there is no information on the yield and size of locked-in capital.  In Grubel (2000), I report on these and other considerations.  Suffice it note here that the lock-in effect undoubtedly has a negative effect on the productivity of Canada's overall stock of capital and that it can be blamed for the empirical estimates of the cost of the capital gains tax presented below.

The Marginal Efficiency Cost of Taxes

The ideas that taxes cause costly distortions and reduced supplies of the taxed factor of production have been formalized in the theory of the marginal efficiency cost of taxation.

This theory argues that a consumption tax tends to have the lowest efficiency cost since every person needs to buy goods and services to maintain desired living standards.  The highest efficiency costs in turn arise from the taxation of capital, for reasons just discussed.  It is easy for wealthholders not to save and invest but consume their income; to move their capital abroad or to keep it locked up in less than optimal investments.

Estimates made by the Department of Finance in Canada and published by the OECD (1997) suggest that the efficiency cost of taxes on capital is $1.55.  That is, for every dollar raised, there is a loss of efficiency equal to $1.55.  Put differently, if a private person has an investment yielding 10 percent and the government taxes it away, society will be poorer unless the government invests the money in a project yielding 25.5 percent.

The most important implication of the marginal efficiency cost models is that there exists an optimal mix of taxes because, according to the Department of Finance study, the efficiency cost of sales and personal income taxes to be only $.17 and $.56, respectively.  These estimates suggest that Canada's tax system could be improved greatly by the elimination of taxes on capital and their replacement by sales and personal taxes.1

Quantitative Estimates of the Cost

The preceding conclusions can be strengthened by considering studies of the size of the output losses caused by capital gains taxation.  Such empirical studies are very difficult because of the myriad of confounding changes in policies, technology, terms of trade and demographics taking place at the same time changes in capital gains taxes are implemented.  Moreover, there are few distinct, major changes in capital gains taxation levels and structure.  And, as noted above, the effects of the tax on the capital stock and productivity are likely to take place over a prolonged period of time.

However, there are two sets of empirical information, which shed important light on the cost of the capital gains tax.  The first was produced by Kugler and Lenz (2001) for a Fraser Institute conference held in Vancouver last year.2   This landmark study is extremely simple in its conception and uses a unique set of data, which are almost like those produced by a laboratory in the natural sciences.

The federal government of Switzerland does not have a capital gains tax.  However, during the last 20 years, individual cantons have experimented with the tax.  Kugler and Lenz showed that cantonal incomes between 1978 and 1995 followed a very similar trend since the same external factors and federal policies affected the economies of all.  However, 8 of the cantons during the period under study eliminated the capital gains tax in their jurisdiction while the rest did not.  Regression analysis showed that the cantons, which eliminated the tax, enjoyed a jump in cantonal income of 3.2 percent relative to the income of those that retained the tax.  The time period under study is too short to make reliable estimates of the effect on the trend rate of economic growth caused by the elimination of the tax.

One obvious question about these empirical results is the extent to which the gains of the cantons that abolished the tax were made at the expense of those that had not.  The authors tested for this hypothesis and found no evidence that the gains of some cantons came at the expense of others.

Undoubtedly, the study by Kugler and Lenz will receive careful scholarly scrutiny and the precise estimate of the effect may change.  However, given the strong theoretical basis underlying the specification of the model, there is little doubt in my mind that the authors' general conclusion will stand up:

"The data ... suggest that the abolishment of the capital gains tax has had positive and economically significant effects on the level of real income in the cantons." (Kugler and Lenz (2001), p.xx).

          The second set of empirical information is based on studies concerned with the much broader issue of the economic benefits of replacing indirect with direct taxes, or more specifically taxes on income with taxes on consumption.  These studies draw on the experience of industrial countries, which have significant differences in taxation systems and reliable data.  Dahlby (2001) in his contribution to this conference presents a careful review of the results of studies using this database.  He concludes:

"...recent econometric studies indicate that switching the tax mix toward consumption taxes can significantly increase the growth rate of the economy.  An increase in the annual economic growth rate on the order of a two-tenths of a percentage point...would have a dramatic impact on Canadian living standards over time.  The present value of the additional per capita output from increasing the annual growth rate by two-tenths of a percentage point is 238 percent of current output, which represents an enormous gain...even if we have overstated the growth effect...by a factor of 10... the present value of the increase in per capita output would still be 22.3 percent of current output – a very significant gain."  (Dahlby (2001), p. x))

Dahlby's conclusion draws on the findings of studies, which considered the entire range of indirect taxes, mainly taxes on labour and capital.  For this reason, the empirical estimate quoted is not entirely relevant to the capital gains tax alone, which produces only a small fraction of the entire revenue from indirect taxes.  According to one estimate, the capital gains tax revenue of the federal government in 1992 represented only .3 percent of all federal taxes collected (Grubel (2000), p. 14).  Poddar and English estimated that federal capital gains taxes represented only 21.1 percent of the total federal tax on investment collected (Poddar and English (1999).  On the other hand, the marginal efficiency of capital taxes generally and specifically the capital gains tax is higher than that for other indirect taxes, which implies that its replacement with a direct tax would have a disproportionately large effect, albeit on a small base.

          Finally, I made a very simple calculation of the rate of economic growth experienced in the period 1990-97 by OECD countries that do not have any capital gains tax imposed by their federal government.   I found that Hong Kong, the Netherlands, New Zealand, Singapore and Switzerland – countries without the tax, had per capita growth rates of 2.2 percent while the rest of the OECD countries grew at 1.2 percent.3 

Conclusions

Theoretical considerations suggest that the efficiency cost of capital gains taxation takes the form of a lower capital stock and its less efficient use through the lock-in effect and the distortions caused by efforts to avoid the tax.  All of these effects translate into lower productivity, per capita incomes and living standards.

These theoretical considerations in turn have been used to make econometric estimates of the quantitative effect on income.  The studies noted above are not specific enough or sufficient in quantity to allow me to conclude that the elimination of the capital gains tax in Canada would result in a certain percentage increase in per capita incomes.  However, together, the empirical evidence implies strongly that the capital gains tax has reduced Canadian income by quantitatively significant amounts and that it will continue to do so in the future.

II.      The Effects on Equity

Many Canadians believe that a necessary part of a desirable, good society is a fair tax system and a fair distribution of income.  The concept of fairness is value-laden and people are entitled to their own views on this subject.  However, taking the pulse of public opinion – and possibly trying to provide leadership – the political establishment decided in 1972 that a capital gains tax in Canada would increase the fairness of the tax system according to norms accepted by the majority of Canadians.

Fairness of the System – "a buck is a buck"

The public mood in the late 1960s and early 1970s was made receptive to the imposition of a capital gains tax on the grounds that in its absence the tax system is unfair in the sense that people with the same level of income pay different amounts of taxes.  In economics jargon, the tax system did not meet the criterion of horizontal equity.  The facts are quite straightforward.  People who have capital gains during a period can increase their spending and living standards (or add to their net worth) to the same degree as if they had obtained the same amount of money in the form of wages, salaries or dividends.  Since the latter types of income are taxed, capital gains should also be taxed.

          The issue of horizontal equity was popularized by the catchy slogan "a buck is a buck".  Its popular appeal was strengthened by reference to the work of two serious scholars, Robert Haig (1921) and Henry Simons (1938).  The so-called "Haig-Simons conditions continue to influence academic discussions about capital gains taxation.

          The basic case for the taxation of capital gains on the grounds of horizontal equity is logically valid as far as it goes.  However, for a number of reasons, it rests on an incomplete understanding of the nature of capital gains.

          First, in a scholarly paper Bruce Bartlett (2001) argues that capital gains are not like income at all.  He supports his detailed arguments by reference to the important fact that capital gains are not counted as income in national accounts, which record the value of goods and services produced.  He quotes Simon Kuznets, one of the creators of the national income accounts "Capital gains and losses are not increments to or drafts upon the heap of goods produced by the economic system for consumption or for stock destined for future use, and hence they should be excluded." (Kuznets (1941), p. 12).

          Bartlett concludes that Haig and Simons did not address the issue raised by Kuznets, if capital gains are not considered to add to national income, how can they considered to add to personal, taxable income?

"In truth, their (Haig and Simons) rationalization for taxation of capital gains rested more on ideological grounds than scientific analysis.  As Haig once put it, 'an income tax which would allow capital gains to escape unscathed would, in this country at least, be an ethical monstrosity.'  Simons too the same basic view.  'The main and decisive case for inclusion of capital gains rests on the fact that equity among individuals is impossible under an income tax which disregards such items of gain and loss.'"  Bartlett (2001), p. 12-13.4

Second, a large proportion of capital gains are illusory increases in spending power of their owners since they are caused by inflation.  Thus, between 1972 and 1991 consumer prices rose 3.8 times while investments in the TSE increased only 2.9 times.  Therefore, capital gains reported on such TSE investments were entirely nominal.  Economic developments did not allow the owners of such capital gains to increase their spending as if the money had come from work.  Their net worth was reduced.5

          It is argued that the system of capital gains taxation in Canada takes account of inflation by requiring the inclusion in taxable income only one half of the realized capital gains.  The preceding data on stock market investments over nearly 20 years show that this provision is not adequate.  The exclusion from taxation of 50 percent of the nominal gains still meant that taxes were due on real losses.  In addition, the inclusion rate introduces a bias in the rate at which investments are taxed depending on the length of time over which they have been held.  Thus, when inflation is moderate, the 50 percent exemption is likely to more than compensate for capital gains held for a short period.  On the other hand, for long-term investments, the exemption is likely to equal the actual rate of inflation, as was the case in the stock market example given in the preceding paragraph.6

          Third, real capital gains tend to be a return to risk-taking by investors willing to place their funds in new ventures.  Such new entrepreneurial ventures are very important for technological progress and for raising the productivity of investment.  Unfortunately, such innovative activities often fail.  Therefore, to attract investors, those that are successful, have to be rewarded with high rates of return.  In the case of many of the new information technology firms, in fact, most of the return to the entrepreneurs have tended to take the form of capital gains on shares they owned when the firm went public.

In principle, the ability to offset capital losses against capital gains in the determination of taxes owed.  In practice, this principle works with venture capitalists with enough resources to have portfolios with stakes in many start-up enterprises.  However, most individual investors are not in such a position and the capital gains tax reduces the expected and realized returns to risk-taking.

The lower the return to an activity, such as risk-taking associated with innovation, the less of it is supplied.  As a result, economic growth and the living standards of Canadians are lowered correspondingly.  The reduced risk taking due to the capital gains tax in effect is one of the reasons why the marginal efficiency cost of capital gains taxation is so high.  In this important sense, the slogan "a buck is a buck" is incorrect.  A buck earned through capital gains has more detrimental effects on productivity growth than does a buck earned through wages.

A corollary of the horizontal equity argument is that creative bookkeeping and tax planning make it possible for people to shift taxable personal and investment income into non-taxable capital gains.  To the extent that this opportunity is exercised, Canada's system of taxation is even less fair than it would be otherwise since it allows some Canadians to escape tax payment altogether while the resources at their disposal, if earned as a salary, would draw large tax payments.

          As noted above, this argument has been used by proponents of the capital gains tax during the period leading up to its introduction in 1972.  At the 2000 Fraser Institute symposium on capital gains taxation, the empirical importance of such income shifting was discussed in the context of the experience of a number of countries that do not have capital gains taxes.  This subject deserves a longer discussion and is found in below in Part IV.

Vertical Equity – "the rich get richer..."

Public opinion in Canada and most other countries is very much concerned with the fairness of the distribution of income and wealth.  They believe that in a good society the poor are raised out of poverty by government programs and transfers financed out of taxes paid by high income earners.  In particular, the public considers it to be important that government policies prevent "the rich from getting richer and the poor from getting poorer", a slogan that is used frequently and effectively by the proponents for what economists have called "vertical equity".  Politicians have responded to these popular demands for greater vertical equity by the creation of progressive income taxes, the double taxation of income from property and the provision of a wide and generous social security program.

          These income redistribution policies give rise to many questions of principle and fact, which cannot be discussed here.  Suffice it to note that the top 10 percent of all income earners in Canada have traditionally paid around 45 percent of all income taxes.  Is it really fair to add to this high burden capital gains taxes, which further add to this burden of this top 10 percent of the income earners?7

          Another problem associated with the redistribution policy is that it does not so much reallocate income between individuals and families as it does for individuals through time.  This is so because people go through life cycles of income.  They have relatively low incomes when young and without work experience.  They have high incomes during the middle and late years before retirement.  When they have stopped working, they tend to have again relatively low incomes (and relatively large assets out of which they finance their retirement).

It is possible to simulate the income distribution, which results under the assumption that there are several generations of people at different phases of their life cycle.  Drawing on actual incomes of Canadians in the different age groups, Sarlo (1997) reports that perfect equality of income of Canadians during their lifetimes results in a static income distribution, which closely resembles that observed in Canada.  In other words, a very large proportion of the income inequality that allegedly requires strong equalization policies by government is due to nothing more than the fact that people tend to have relatively low incomes when they are young and old.  Capital gains taxation does not equalize across the size distribution of income as much as it forces equalization of incomes through life.

          A third reason to question the validity of the argument in favor of income redistribution was presented to me a number of times by Canadians during private discussions and following my presentation of the economics of capital gains taxation.  These Canadians asked "What is fair about taking away from some the rewards of their work, investment and risk-taking and transferring it to those who have worked less hard, saved and invested less and took fewer risks".

Answers to this question are very divisive since, of course, there are also many Canadians who have worked hard, saved and did all the right things but fate intervened and they have low incomes because of illness, accidents or simply bad luck.  Most Canadians believe that such people deserve a helping hand from society.  The basic problem associated with all policies of income and wealth redistribution is the difficulties associated with identifying those who deserve support because their condition is due to forces beyond their control and those who are responsible for their state of need as a result of their own decisions.

It is important to raise these considerations about the causes of need in the present context where the case of the capital gains tax rests to some degree on the demand for greater income equality through government policies.  Politicians should be aware that a growing number of Canadians are concerned about this problem.  They should also lead a discussion on the role that private charity should play in alleviating the conditions of those with low incomes.

Is the Tax on the Rich?

The taxation of capital gains is considered to be highly desirable by the proponents of greater vertical equity in the distribution of income because of statistics, which show that the rich pay most of the capital gains taxes.

          Calculations made by Joel Emes of the Fraser Institute8, using a data base supplied by Revenue Canada found the following: in 1992 families with income of $100,000 and more paid 78 percent of all capital gains taxes.  Families with income of $50,000 or less paid only 8 percent of the total.  In looking at these numbers, it should be remembered that only 7.9 percent of all Canadian families in 1992 had incomes above $100,000.

          However, the validity of these data has been questioned on the basis of the fact that the income used to classify families includes capital gains.  As a result, it is possible that the statistics on the high-income earners are misleading.  These apparently rich Canadians may have much lower incomes in years other then the one in which they report their capital gains and pay the tax.

          Emes examined this issue employing the same 1992 data he had used to produce the conventional statistics on the incidence of the capital gains tax just discussed.  He found that those Canadian families with incomes other than capital gains over $100,000 in 1992 paid only 26.8 percent of all such taxes.  Families with less than $50,000 of such non-capital gain incomes paid 52.1 percent.

          These results can be explained most easily in the context of the normal life cycle of incomes.  The owners of a small business often take out little income annually and reinvest much of their profits to expand their business, keep up with the competition and, most important, to build up a nest-egg for their retirement.  When these owners of small businesses sell it to retire, they appear as "rich" Canadians in the statistics of that year.  In fact, of course, their incomes both before and after the event often do not make them the kind of high-income earners at whom the capital gains tax is aimed in order to create greater vertical equity.9

          In addition, during the last few decades Canadians from all segments of the income distribution have increasingly invested savings in mutual funds.  Under present rules, when a mutual fund sells assets and makes capital gains, they give rise to capital gains tax obligations by the individual owners of the mutual funds.  As a result, many Canadians with low and modest incomes have had the personal, unpleasant surprise of facing unanticipated tax obligations when they file their annual returns.  Some have had to sell part of their fund holdings to generate the cash needed to pay the tax collector.  These reactions of wealthholders to the capital gains tax induce an inefficient turnover in capital and contribute to the marginal efficiency cost of taxes on capital noted above.

The Incidence of the Tax

Economists tend to stress the fact that a tax on automobiles does not depress the return on the capital earned in the industry.  In other words, the big automobile companies are not paying the tax because, as analyzed above, a small country like Canada needs to maintain a competitive after-tax rate of return on capital.  Instead, a tax on automobiles leads to higher prices, which are paid by all Canadians, who buy cars.  It is therefore largely illusory to believe that a tax on cars leads to lower returns to the owners of capital and the greater equality of after-tax income demanded by many.

          There is a strict analogy between the analysis of the incidence of a sales tax on automobiles and any tax on capital employed in the automobile and other industries in Canada.  As noted above, the imposition of a capital gains tax leads to a rise in the before-tax rate of return to assure that the after-tax rate in Canada remains the same as that prevailing in the rest of the world. The higher after-tax returns lead to a lower capital stock per worker and therefore lower wages.  Labor, not the owners of capital end up paying the capital gains tax.

It is important to remember this argument in the present context of the analysis of equity issues, especially since this burden on labor falls more on those with low incomes who spend a much larger share of their incomes than the rich on goods and services.  Therefore, it is largely illusory to believe that the capital gains tax results in a greater equality of income.10

Conclusions

The capital gains tax was introduced on the basis that it would increase the fairness of the Canadian tax system and income distribution.  In this section, I have questioned the conventional wisdom that "a buck is a buck" should be the basis for an equitable tax system.  Basic economic theory suggests that capital is not like income when it comes to national income accounting.  Taxing the fruit and the tree has important implications for the size of future fruit harvests.  Many capital gains are due to inflation and the capital gains tax often amounts to the confiscation of real capital holdings. Some capital gains are the return to risk-taking.  Taxing this activity leads to a reduction in this activity, which is very important for economic growth.

          The capital gains tax is also designed to equalize the distribution of after-tax income – "to prevent the rich from getting richer and the poor from getting poorer".   The rationale for this policy is based on the conventional wisdom that mainly high-income earners with incomes of more than $100,000 a year pay the capital gains tax.  In fact, however, families that earn less than $50,000 in income other than capital gains pay over half of the tax.  In addition, in a small country like Canada, capital taxes do not result in lower after-tax returns on capital.  The cost of the tax is shifted to workers through lower wages and to the buyers of products through higher prices.  Such shifting makes largely illusory the alleged beneficial effects of the capital gains tax on the after-tax distribution of income in Canada.

          In the light of the preceding analysis we may conclude that there are few real beneficial effects brought by the capital gains tax for the fairness of the tax system and the distribution of income.  Combining the findings reached in part one about the high efficiency cost of the tax with the findings about its minimal effect on fairness suggest that Canadians would be better off if the tax were abolished.

III.     Revenue Implications

One of the justifications for the imposition of the capital gains tax in Canada in the early 1970s was to raise revenue to finance the general operations of government, which in the spirit of the times was at a strong upward trend.  There is considerable evidence that the tax may have achieved this objective in the short run but that its detrimental effect on economic growth has resulted in a reduction in overall tax revenues.

          As already noted, it is difficult to study the longer run effects of the capital gains tax on economic growth and the accompanying tax revenues.  However, it is much easier to find short-run effects, especially in the United States, where capital gains tax rates have been changed a number of times during the last 50 years.  Revenue data show clearly that decreases in the rate resulted in higher revenues and that revenues declined in anticipation and in the wake of rate increases.11  This result is not difficult to explain.  Since the realization of capital gains is at the discretion of the owner of the capital, so is the payment of the tax.  Owners of capital manage their affairs through time to minimize their tax obligations.

          Questions have been raised about the validity of these findings in the longer run.  On a logical basis wealthholders with finite lives cannot postpone the realization of capital gains forever.  At some point, even if only after death, capital gains taxes are due.   For this reason, lower rates will results in lower revenues.  Of course, this argument applies with extra force when the capital gains tax is eliminated completely.  In that case even in the short run revenues collected from the tax will be zero.

          The preceding analysis is deficient because it focuses too narrowly on the relationship between rates and revenues on the capital gains tax alone.  As argued above, reductions in the rate or the elimination of the tax stimulate longer run growth in productivity and income through the deepening of the capital stock, its more efficient use and other induced changes.  Such higher income results in greater revenues from personal, sales and business taxes.  The question is whether the revenue gains from these taxes are likely to be greater than the losses from the lower rate or the elimination of the capital gains tax.

          To answer this question we must look at the revenue presently raised through the capital gains tax.  In this context it is interesting to note that published government documents – the annual Budgets, publications by Statistics Canada or the Department of Finance – do not contain this information.  OECD statistics show capital gains tax revenues for most industrial countries, but not Canada.12  For this reason, Joel Emes at the Fraser Institute produced an estimate from a Revenue Canada database.  As noted above, he estimated the 1992 revenue to be $715.6 million, or about .3 percent of the total tax revenue that year.  Even if this estimate is off by a factor of 3, the total revenue is quite small.

          The other side of the coin is the expected gain in general tax revenues due to the higher level of income stimulated by the elimination of the tax.  Section one above discussed the empirical evidence produced by Kugler and Lenz on the effect on the level of income in some Swiss cantons resulting from their elimination of the capital gains tax.  The section also presented the empirical evidence reviewed by Dahlby on the efficiency and income effects of the elimination of taxes on capital.  Both of these papers suggest that the elimination of the capital gains tax would result in a significantly higher level of income.  We may therefore expect also a corresponding increase in general tax revenue.

          Unfortunately, the available empirical evidence is not strong enough to permit the conclusion that the elimination of the capital gains tax will result in overall higher tax revenues.  However, the evidence is strong enough to conclude with confidence that the overall revenue loss would be very small at worst and that there would be significant gains at best.  Under either scenario, revenue considerations should not play an important role in the decision to eliminate the capital gains tax in Canada.

IV.     The Costs of Administration

When a business has to pay a tax on profits but not on capital gains, incentives are created to use creative bookkeeping and legal maneuvers to turn profits into capital gains.  The process of achieving this goal has become known as "surplus stripping". 

Essentially the process requires that the owners of a working company establish a dummy corporation.  This dummy borrows money from a bank to buy all of the shares of a working company.  The dummy company becomes the owner of all of the real assets of the working company plus the profits it earned in the preceding period.  These profits are referred to as surplus and are to be turned into capital gains.

The next step in the process of stripping this surplus is for the working company to use the money it received from the dummy corporation to buy back all of its shares.  It thus becomes again the owner of its real assets but by agreement, it leaves the surplus with the dummy corporation.  The dummy then uses the money it received for the shares to repay the bank loan.  However, it has left over the surplus, or profits of the working corporation.  The dummy corporation is dissolved and in the process declares a capital gain equal to the working company's profits.  The owners of the dummy corporation in effect have turned the taxable profits of their working company into a non-taxable capital gain.13

A publication by the Department of Finance (1980) reviewed the history of problems, which led to the adoption of the capital gains tax.  It reached the following conclusion:

"The inability of the government to check surplus stripping abuses was, in fact, the primary impetus for the comprehensive review of the tax system in the early 1960s.  It led to the establishment of the Royal Commission on Taxation."

The Carter Commission Report recommended the adoption of the capital gains tax to prevent surplus stripping.  The government accepted this recommendation and adopted a capital gains tax in 1972.

          At the 1999 symposium on capital gains taxation at the Fraser Institute, a number of participants, who had been working in the financial sector during the 1960s disputed the proposition that surplus stripping was a major problem.  They noted that such stripping is not possible for large, publicly held companies.  Only small, privately held companies were able to engage in this kind of legal and financial maneuver.  It was not difficult for the Department of Finance to prohibit such maneuvers, which so obviously had no other purpose than the avoidance of taxes.  These observers suggested that surplus stripping was a red herring to divert public attention from what essentially was an increase in the overall burden of taxation to finance increases in the size of government and its ability to redistribute income.

          To shed further light on this issue, the 2000 symposium brought to Vancouver economists who were knowledgeable about the problems experienced in their countries as a result of having no capital gains taxation.  Two of the papers are of special relevance to the present analysis.

Empirical Evidence from Hong Kong and New Zealand

Berry Hsu and Chi-Wa Yuen (2000) discussed Hong Kong's experience.  They provided a list of methods that can be used to turn profits into capital gains.  Some are quite simple, like buying real estate with the intention of reselling it, but claiming that the resulting profits were capital gains.  Others are more sophisticated, like the sale of the right to a future stream of income, claiming it to be a capital gain.

          These and other schemes have been tried and have been challenged successfully by the Hong Kong revenue authorities.  These authorities are guided in their actions by two, relatively simple anti-avoidance principles:

"The first sets out to disregard any 'artificial or fictitious' transactions that do not in fact take place and any that reduce or would reduce the amount of taxes payable.  The second applies a 'sole or dominant purpose' test to determine whether a transaction is conducted mainly for the purpose of obtaining tax benefits through the avoidance or postponement of the liability to pay tax or the reduction in the amount thereof." (Hsu and Yuen (2001), p. xx)"

The decisions of the revenue authorities based on these principles are challenged regularly by the affected parties.  A Board of Review hears adjudicates these appeals.  Hsu and Yuen consider the rulings by the revenue authorities and of the review board to reflect the magnitude of the efforts of taxpayers to shift taxable income into non-taxable capital gains.  They conclude:

"On the basis of the indirect evidence available to us we conclude that the absence of a capital gains tax in Hong Kong has resulted in little, if any, inefficiencies and inequities."  (p. yy)

Hong Kong, in so many ways is like no other "country" in the world and its experiences may have only limited relevance to others.  Most important, the low income and profits taxes, which keep low the returns to turning taxable income into non-taxable capital gains.

            On the other hand, Hong Kong's practices for dealing with such tax avoidance efforts show that it should be possible for Canada's revenue agency to set out some simple anti-avoidance principles against which it, and an appeals court, can evaluate any practices that appear to be designed to avoid taxes.  Such a task would not be new since the present more favorable rate of taxation of capital gains relative to other income already contains strong incentives to avoid taxes by dividend stripping and the methods used in Hong Kong.  Existing rules and regulations may be expected to apply fully in a regime completely without capital gains taxes, though of course, the incentives to get around these rules and regulations would be stronger for the private sector.

          The paper on New Zealand by Robin Oliver (2001) gives a hint of the kinds of difficulties encountered in a country that is more like Canada than Hong Kong is.  Oliver, as a former tax consultant and in his present position as an adviser in the revenue department, brings a point of view shaped by his day-to-day battles with the private sector, which is constantly inventing new methods for avoiding taxes through the zero capital gains tax loophole.  He concludes:

"I would assert, after my long experience with the New Zealand tax regime as a private consultant and government adviser, that the best possible system is not one, which simply excludes capital gains from taxable income." (Oliver (2001), p.xx)

In the concluding section of his paper, Oliver modifies this strong statement with the following:

"This paper has canvassed the problems posed for New Zealand's tax system by the absence of a general capital gains tax.  Undoubtedly, if we had a capital gains tax, the paper would have canvassed the problems posed by having such a tax." (p.yy)

Conclusion

The study of the experience of countries without capital gains taxes suggests that the abolition of that tax in Canada would increase the troubles and cost of preventing tax avoidance incurred by the revenue authorities.  It would also increase private expenditures on efforts to find loopholes for avoiding the tax and staying ahead of the tax authorities' efforts to close those already in use.  On the other hand, the experience of other countries also suggests that the task is manageable and can be handled by a large bureaucracy already dedicated to limiting tax avoidance and evasion.  However, there is no doubt that the expected higher costs for the prevention of avoidance should enter into the benefit/cost analysis needed to make the case for the elimination of the capital gains tax.

V.      Summary and Policy Implications

At the risk of oversimplification, the battle over the retention or abolition of the capital gains tax is fought along lines characterized by two catchy slogans:  "A buck is a buck" and "Tax the fruit but not the tree".  The first slogan is about the distribution of income and the fairness of the taxation system.  The second is about the efficiency implications of the tax and how much it reduces economic growth and per capita incomes.

          Most people agree that both fairness and efficiency should be considered in the design and operation of Canada's tax system.  The crucial issue arises around the magnitudes of these two social objectives.

          This paper discussed the growing evidence on the size of the efficiency losses due to the tax.  The large number of studies using the concept of the marginal efficiency cost of taxes implies that the cost is substantial.  The study of the Swiss experience is powerful and adds to direct evidence on the effects of thet tax on income.  These results support the notion that capital gains are not like other income from a number of points of view that are important from an economic point of view.

          The prediction of higher income levels in the wake of the elimination of the tax imply that tax revenues from all sources are likely to increase if the capital gains tax is eliminated.  On the issue of vertical equity, recent calculations have shown that the incidence of the capital gains tax is much less progressive than had been believed.  Over half of it is paid by families with incomes other than capital gains, which most people would consider to be moderate rather than high.  On the issue of horizontal equity, the evidence from Hong Kong and New Zealand suggests that revenue authorities are able effectively to limit attempts at shifting taxable income into non-taxable capital gains.

          The quality of the empirical evidence available is not sufficient to engage in a formal benefit/cost analysis of the elimination of the capital gains tax.  However, in my view the evidence is strong and clear enough to suggest that the elimination of the tax will improve the incomes and well being of Canadians by a margin large enough to outweigh the expected costs of administration and in the form of reduced fairness.

References

Bartlett, Bruce (2001), "Why the Capital Gains Tax Rate Should be Zero". Dallas, Texas: NCPA Policy Report No. 245, August

Dahlby, Bev (2001), "Restructuring the Canadian Tax System by changing the Direct/Indirect Tax Mix", Paper for Fraser Institute Conference on Taxation, Toronto, Ont., October 11, 2001

Department of Finance (1980), " A Review of the Taxation of Capital Gains in Canada: An examination of the Canadian experience and of issues involved in proposals for change", November

Eisner, Robert (1980), "Capital Gains and Income: Real Changes in the Value of Capital in the United States, 1946-77", in Dan Usher, ed., Measurement of Capital, Chicago: University of Chicago Press, 1980

Grubel, Herbert (2000), Unlocking Canadian Capital:  The Case for Capital Gains Tax Reform, Vancouver: The Fraser Institute

------------------- (2001), International Evidence on the Effects of Zero Capital Gains Taxes, Vancouver: The Fraser Institute

Hsu, Berry F. C. and Chi-Wa Yuen (2001), "Tax Avoidance due to the Zero Capital Gains Tax: Some Indirect Evidence from Hong Kong", in Grubel (2001)

Haig, Robert M. (1921), "The Concept of Income – Economic and Legal Aspects", in Robert M. Haig, ed., The Federal Income Tax, New York: Columbia University Press

Jorgensen, Dale and Kun-Young Yun (1991), "The Excess Burden of Taxation in the United States", Journal of Accounting,  Auditing and Finance, 6(4): 487-509

Kugler, Peter and Carlos Lenz (2001), "Capital Gains Taxation: Some Experience from Switzerland", in Grubel (2001)

Kuznets, Simon (1941), National Income and Its Composition, 1919-38, New York: National Bureau of Economic Research, 1941

OECD, (1997), Economic Surveys, Canada, Paris, page 85

Oliver, Robin (2001), "Capital Gains Tax: The New Zealand Case", in Grubel (2001)

Poddar, Satya and M. English (1999) "Canadian Taxation of Personal Investment Income", Canadian Tax Journal, 47,5,1270-1304

Simons, Henry C. (1938), Personal Income Taxation, Chicago: University of Chicago Press

Notes

[Contents]

info@fraserinstitute.ca
4th Floor, 1770 Burrard Street, Vancouver, BC, Canada, V6J 3G7
Tel: (604) 688-0221 Fax: (604) 688-8539 Book Orders: 1-800-665-3558 ext. 580

You can contact us at the above email address for any comments or information requests. Please report any dead links or technical problems.