A CLOSER LOOK AT THE FLAT TAX
TABLE OF CONTENTS
A CLOSER LOOK AT THE FLAT TAX
The ultimate in flat tax reforms would create a single and simple system comprised of two basic ingredients:
Flat tax advocates argue that these changes would not only lead to a simpler tax system, but also improve economic efficiency, which means there would be fewer impediments to businesses making decisions for purely economic reasons and individuals making decisions based purely on their preferences. Most real-life flat tax proposals fall short of these goals, imposing a uniform marginal tax rate while retaining many and, in some cases, all of the deductions, credits and tax breaks in the existing system. Hybrid flat tax proposals have been surfacing and resurfacing in the United States since at least the early 1980s. In this country, the Canadian Alliance has done much to bring the issue to the fore, recently proposing its own flat tax but subsequently shelving the idea at least temporarily.(1) Nevertheless, if the United States is any guide, the issue is likely to re-appear in the policy arena, especially given the ever-changing and increasingly complex nature of the taxation system and the strong forces promoting international harmonization of taxes and regulations.
This paper has three objectives:
Before answering these questions, it is useful to briefly review two key concepts of tax equity from the theory of public finance.
Other terms including progressive, proportional and regressive tax regimes need to be defined because they turn up frequently in popular discussions and can easily be abused or confused. Most economists define a progressive tax regime as one where those with lower income pay a smaller share of their total income in taxes than those with higher income. A regressive taxation regime is just the opposite: lower-income persons or families pay a greater share of their income in taxes than do higher-income persons or families. A proportional tax regime is one where the same marginal tax rate applies to all persons, regardless of income, so that everyone pays the same proportion of their income in taxes.
In the extreme, a regressive tax would mean that low-income earners would pay a higher tax rate than high-income earners. Regressive taxes, however, are often a lot more commonplace than is generally thought to be the case. Taxes on food, for example, are the same for everyone and yet are considered regressive because food purchases account for a bigger portion of total income for a typical low-income individual or family than they do for their high-income counterparts. The following table helps to illustrate the differences between each type of tax.
Table 1: Comparing Different Tax Régimes
In the popular debate, there is also much confusion about whether a flat tax, which is proportional in a strict definitional sense, can be effectively progressive, i.e., can it assure that a low-income family or individual pays less of its total income in taxes than does a high-income family or individual? Given a sufficiently large basic tax exemption, the answer is an unambiguous yes.(3) To see why, consider the table below, where we assume two families (the Smiths and Simpsons, each with a sole breadwinner) and compare a progressive tax regime with a proportional tax regime.
Table 2: Comparing Progressivity
The first column shows a hypothetical graduated or progressive tax system (with a $10,000 basic exemption) that imposes a marginal tax rate of 17% on the first $100,000 of income and 25% for income above that threshold. Because the Smith familys income exceeds $100,000, the Smiths pay both tax rates for a total tax bill of $27,000, or 18% of its income. The Simpsons, on the other hand, pay only $5,100 or 12.8% of their income. This system is progressive on two counts. First, in the formally progressive sense defined by William Vickrey, namely because it imposes different marginal tax rates for different income levels (17% and 25%) and second, because the average tax rate (calculated after factoring in basic exemptions) is lower for low-income families. The fact that the Smiths pay more in dollar terms than the Simpsons says little about whether the system is truly progressive, at least given the way most economists have defined the term. The Smiths would pay more than the Simpsons even under an extremely regressive (and improbable) tax system that imposed a 4% tax rate on all the income of high-income families, but a 17% tax rate on all the income of low-income families. In this extreme case, the Smiths would pay $5,600 versus $5,100 paid by the Simpsons.
Now suppose theres a sudden switch to a flat tax system. The second column shows how a flat or proportional tax given a large enough basic exemption (for illustrative purposes, the exemption has been increased to $20,000 from $10,000 in the progressive regime) can be at least as progressive as, if not more progressive than, a formally progressive tax.(4) Under this scheme, the Smith familys effective tax rate falls to 14.7% and it saves a net $4,900 relative to the progressive regime, while the Simpson familys effective tax rate falls to 8.5% and it saves a net $1,700 relative to the progressive tax system. Again, it must be stressed that just because the Smiths save more than the Simpsons under this flat tax proposal does not mean the system is regressive. The Smiths still pay a greater portion of their total income in taxes than do the Simpsons. It would be hard to avoid this outcome (i.e., a greater absolute tax saving for the high-income family) given any meaningful reduction in marginal tax rates. Generally speaking, the most effective way to give more to lower-income families without an even bigger saving for the wealthy is to have targeted deductions and exemptions.
Finally, column three shows what could be called a doubly pure flat tax because it has no basic deduction. Under this system, the Simpsons are unambiguously worse off than under the progressive system or the proportional system with large exemptions. Meanwhile, the Smiths still fare better than they do under the progressive system but not quite as well as under the flat tax with a large basic deduction. This analysis illustrates two key points:
The term flat tax, at least insofar as it has gained widespread currency, has been around since 1981, when the idea was put into the policy arena by Robert Hall and Alvin Rabushka. These two Stanford academics published a book entitled The Flat Tax containing the essential features of their proposal which can be summed up in five points.
Hall and Rabushka argue their plan would not only lead to a simpler tax system tax forms for most citizens and businesses would be half a page long, at most but also promote efficiency and hence economic growth in the long run. The current system, they argue, is too complex and poorly designed, leading to unanticipated and negative consequences. For example, U.S. employers are increasingly paying their employees in fringe benefits in lieu of wages or salary because they can deduct these benefits as expenses. At the same time, employees receive the benefits tax-free. A flat tax system would treat these kinds of compensation equally and lead to higher pay for employees who would then have more freedom to make decisions in their own best interest (about, for example, life insurance) rather than have them imposed by an employer whose motivation may be purely tax-driven. The economy, as a whole, would consequently enjoy efficiency gains.
Similarly, Hall and Rabushka argue their one-year write-off provision would increase investment because firms would have a strong incentive to add to their physical capital (machines, buildings, factories) because of the tax savings. The two academics also argue that investment would increase because their flat tax proposal is a true consumption tax. To the extent the economy is constrained by an inadequate supply of savings, a flat tax would improve the economys growth prospects by encouraging saving, which would lower interest rates and make investment in physical capital more attractive.
By the mid-1990s, at least eight flat tax proposals most of them hybrids of the Hall-Rabushka proposal and the existing tax system were on the table in the United States. Many of these had been put forward by Republican presidential candidates but some also came from Democratic representatives (Banks, 1996). The bipartisan nature of these proposals speaks to the fact that a flat tax has the potential of being fair to low-income families and individuals provided it features some kind of large minimum deduction. Most of the proposals followed the Hall and Rabushka plan in spirit but did not eliminate key deductions, one of the most important being the deductibility of interest payments on home mortgages, so cherished by Americans; the construction industry lobby had a strong interest in retaining this feature of the tax code. The end result was that most flat tax proposals were not revenue-neutral: absent draconian spending cuts, they would have implied increased deficits, something that was not politically palatable in the mid-1990s.
However, the idea of a flat tax on income minus the name flat tax dates much further back than the 1980s. In fact, the idea of taxing all personal and business income at the same rate is as old as the income tax itself, at least in the United States. In the late 19th century, for example, proponents of a U.S. proportionate (i.e., flat by another name) tax saw it mainly as a means of redressing regressive tariff and excise taxes, which at the time accounted for the bulk of government revenue.(5) In the context of that era, any kind of income tax, including a flat tax, was seen as a fair means of attenuating the dramatic rise in inequality that accompanied rapid technological change and industrialization in the late 19th century.(6)
In Canada, the federal government began taxing personal income in 1917 to pay for the large expenses incurred for the war effort.(7) Until that point, Ministers of Finance had contrived almost annually in their budget speeches, by the use of most inadequate statistics, to prove to their own satisfaction that Canada was one of the lowest taxed countries in the world (Perry, 1955, p. 144).(8) Although initially seen as only a temporary measure to meet pressing needs, income taxes quickly became a fixture of government fundraising, especially with the twin crises of the Great Depression and the Second World War, both of which helped consolidate the federal governments jurisdiction in this area.(9)
From the beginning, Canadas income tax system was progressive, perhaps due to the fact that many were concerned about war profiteering by industrialists. Although all income was taxed at a so-called normal rate of 4%, graduated surtaxes of 2% (on income of $6,000) to 25% (on income of $100,000 or more) were imposed once certain income thresholds were crossed. The tax system also included basic personal exemptions of $3,000 for married persons and $1,500 for single persons on the normal tax rate (and not the surtaxes), which added another heavy layer of progressivity to the tax system. One estimate suggests that the original income tax system affected at most 1% of the total population.(10)
The original Income War Tax Act was a simple document, taking up all of ten pages. It was little changed until 1962, when the Carter Commission recommended a more comprehensive tax base. After a series of public hearings and a White Paper on Tax Reform, new legislation was introduced in 1972 that modernized the tax act, effectively repealing and replacing virtually all of the old tax laws and forming the backbone of Canadas current system.(11)
Despite the superficial intuitive appeal of the idea and the high-profile nature of the U.S. debate, the flat tax has been adopted by only a handful of countries. The most widely cited historical example is Hong Kong, which imposes a maximum tax rate of 15%.(12) Business groups have cited the countrys flat-tax system as an important part of the Hong Kong economy. There are, however, many other factors that contributed to Hong Kongs strong growth. In fact, it seems unlikely that the flat tax alone or even in large part explains Hong Kongs success because other economies with much different tax structures but similar cultural and geographic characteristics (Japan, for example) have had similar trade records.
The so-called transition economies (i.e., former Eastern Bloc countries) seem to be the most willing to experiment with the flat tax idea. Estonia established a flat tax in 1994 and Latvia followed suit in 1995. In August 2000, Russia adopted a 13% flat tax on income in an effort to meet its revenue needs. Under Russias old tax system, which imposed a graduated tax rate ranging from 12% to 30%, taxation revenue consistently fell short of the governments objective because of tax evasion coupled with the governments inability to enforce its tax laws. Again, it is too early to say and, to our knowledge, no study has shown whether the flat tax had a positive effect on Russias economic growth or government revenue.
The appeal of the flat tax to the transition economies makes some intuitive sense given the historical record. As discussed earlier, the idea of a flat tax first came onto the scene at about the same time as income taxes and that is, arguably, when it had its best chance of being implemented.(13) To the extent that the transition economies are relatively new to capitalism and are starting from scratch, there are far fewer obstacles to imposing a flat tax system. Indeed, many flat-tax proponents in the U.S. and Canada acknowledge that the biggest political obstacle to their plan is the fact that the current tax system is entrenched both at the institutional and personal level, where many persons either make a living off the complexity of the system (tax lawyers and accountants, for example) or are subsidized through the tax system (the construction industry in the United States, for example).
The core of Canadas current tax code has remained remarkably unchanged since 1988. From 1988 through to the beginning of 2001, there were three marginal tax rates of 17%, 26% and 29%.(14) Both the tax rates and the tax thresholds (the levels) were little changed throughout,(15) adding an element of regressivity to an otherwise progressive tax system: individuals whose income rose in line with inflation below 3% often found themselves in a higher tax bracket even though their real situation in terms of income (i.e., after inflation but before taxes) had not changed.(16) Almost by default then, the 2000 budget represents some of the most radical changes to the income tax system since 1988.
Since then, the government has introduced further tax cuts. Beginning in January of 2001, the lowest tax rate fell to 16%, the middle rate was cut to 22%, and income between $60,000 and $100,000 now faces a 26% marginal tax rate instead of 29%. Income in excess of $100,000 will still be taxed at 29%. The capital gains inclusion rate was further reduced to 50%, and the plan to cut all corporate tax rates to 21% (service-sector and high-technology firms generally face higher tax rates the 2000 Budget dropped the tax rate for these firms to 27%) was accelerated.
As with most policy proposals, the flat tax proposal(18) has its advocates and opponents. From a political perspective, flat-tax advocates usually argue a flat tax has three key selling points. They are:
There are a number of less tangible economic arguments in favour of a flat tax, most of which hinge on efficiency considerations. They include:
There are, of course, a number of arguments against a flat tax. Flat tax opponents usually make two visible or popular arguments against the tax:
There are also a number of economic arguments against the flat tax:
Up to now, the discussion in this paper has been rather abstract. This section attempts to present a real-world sense of the budgetary consequences of a flat tax that retains most of the current system of deductions and tax credits. It is important to stress this is very much unlike the Hall and Rabushka proposal that has been examined throughout most of this paper. The rationale for looking at this hybrid flat tax is simply that significant institutional changes are rarely implemented except in times of crisis. This task begins with a question: What is the lowest flat tax that could be imposed if all the features (tax cuts and spending initiatives) implemented in the most recent budget documents (Budget 2000 and the economic supplement) were retained and an assumption was made that future governments are committed to: a) maintaining at least a balanced budget; b) increasing nominal expenditure at the same rate as population growth plus inflation; and c) reducing the debt by at least $6 billion per year? This is clearly a difficult question given the complexity of the tax code. Fortunately, Statistics Canada has developed a model called the Social Policy Simulation Database (SPSD/M) that gives a sense of what a flat tax policy would cost given these conditions. Before looking at the results of this experiment, two important cautionary remarks are in order:
Figures 1 and 2 show results for the calendar years 2000 and 2001.(26) The charts show clearly that a revenue-neutral flat tax given the proposed spending increases, debt reduction plans, and existing system of deductions and credits would have to be somewhere between 19% and 20%. Again, it must be stressed that the figures in this paper were calculated relative to the base case, namely the tax system modified for changes in the 2000 budget as well as the fall economic statement. Note also that the total cost figures include the effects of putting in place a flat tax plus 1% annual spending increases (in line with population growth) and $6 billion in annual debt reduction debt.
The discussion in this paper has attempted to provide a relatively complete contextual picture of the flat tax by looking at its historical roots, its modern-day uses, and arguments for and against the idea. Some likely budgetary consequences of the tax, given the structure of a politically plausible proposal and the current tax system, have also been considered. These calculations should be viewed as merely suggestive. Much can happen in a year or two that would invalidate these figures, potentially making this kind of flat tax more or less feasible. If one conclusion were drawn from this analysis, it might be this: the historical and international record suggest that the odds of moving towards a flat tax increase greatly after an abrupt and dramatic failure of the existing tax system (Russia) or at the earliest stages of an income tax system (Estonia, Latvia and Hong Kong). Short of that, the sheer inertia behind the existing tax system invariably corrupts flat tax proposals beyond recognition, weakening or destroying most of the ideas strongest selling points and ultimately cutting into the revenue-generating potential of the proposal. This in turn threatens core programs, and that may entail high political costs.
Alliance Party. 2000. Solution 17. Internet: http://www.canadianalliance.ca/index_e.cfm.
Beauchesne, Eric. 13 September 2000. Canada Duplicating U.S. Economic Miracle. The Ottawa Citizen, Internet edition: http://www.ottawacitizen.com/.
Canada. 1987. The White Paper on Tax Reform: 1987.
Canada. Department of Finance. 2000. The Budget Plan 2000. Department of Finance Distribution Centre.
Chernick, Howard and Andrew Reschovsky. September-October 2000. Yes! Consumption Taxes are Regressive. Challenge Magazine, pp. 60-91.
Gillespie, Irwin W. 1991. Tax, Borrow & Spend: Financing Federal Spending in Canada: 1867-1990. Ottawa: Carleton University Press Inc.
Hall, Robert E. and Alvin Rabushka. 1995. The Flat Tax. Stanford, California: Hoover Institution Press.
Lerner, S., C.M.A. Clark and W.R. Needham. 1999. Basic Income: Economic Security for All Canadians. Toronto: Between the Lines.
Moore, Stephen. 14 January 1997. The Alternative Maximum Tax. Editorial, Wall Street Journal.
Perry, David P. 1997. Financing the Canadian Federation, 1867-1995: Setting the Stage for Change. Toronto: Canadian Tax Foundation.
Perry, J. Harvey. 1955. Taxes, Tariffs and Subsidies. Toronto: University of Toronto Press.
Rosser, J. Barkley Jr. and Marina Vcherashnaya Rosser. 2001. Inequality and Underground Economies. Challenge Magazine, pp. 39-50.
Slemrod, Joel. 1996. Deconstructing the Income Tax. American Economic Association Papers and Proceedings, pp. 151-155.
Vickrey, William. 1987. Progressive and Regressive Taxation. In Eatwell, John, Murray Milgate and Peter Newman, eds. The New Palgrave: A Dictionary of Economics. London: MacMillan Press.
(1) Flat tax proposals have also been made by Member of Parliament Dennis Mills. As well, the Progressive Conservative Party briefly considered the idea of a flat tax, as did the Reform Party, the predecessor of the Canadian Alliance.
(2) It is, of course, possible to imagine that the no-children family is burdened with large medical costs for one or both spouses or is looking after ill grandparents.
(3) See, for example, Nobel Prize winner William Vickreys definition of progressive and regressive taxation in the New Palgrave dictionary. What he calls formally progressive refers to the idea of different marginal tax rates applying to different income classes. However, he notes that even taxes and tax systems normally considered regressive can be modified to reduce the degree of regressiveness or even render them moderately progressive. An alternative way of achieving a progressive flat tax structure would be to tax all income (no exemptions) equally but combine that system with a large tax credit, essentially putting in place a negative tax or basic guaranteed income provision. See S. Lerner, C.M.A Clark and W.R. Needhams Basic Income: Economic Security for All Canadians for one such proposal. Alternatively, a tax system with adequate exemptions and/or deductions such as the Robert Hall and Alain Rabushka flat tax proposal and some of its variants could achieve a similar blend of horizontal and vertical equity.
(4) This assumes that all else is equal, i.e., that the flat tax doesnt somehow lead to cuts in social welfare programs that transfer money directly to lower-income families.
(5) This was true in Canada as well; see Perry (1997). Tariff and excise taxes were, and to the extent that they still exist are, regressive when they apply to goods and services that take up a greater share of the revenue of a low-income family or individual than of a wealthy family or individual because manufacturers are generally able to pass along the taxes in the form of higher prices. For example, excise taxes on gasoline are generally thought to be regressive because there is evidence that low-income families and individuals spend more of their total income on gasoline than do high-income families or individuals. However, as Gillespie (1991) shows, Canadian officials were historically reluctant to impose excise taxes on basic goods (coffee, tea, food) because of concerns about attracting and keeping immigrants who tended to spend significant portions of their income on these items. Most taxes were applied to so-called sin items such as alcohol. This was made politically feasible because of the forces promoting prohibition.
(6) The first, general income tax was enacted in Great Britain in 1799, to finance the Napoleonic War. The tax was subsequently repealed and re-introduced in the 1880s, when it was generally accepted as a permanent fixture of government finance.
(7) Income taxes were levied by some provinces before World War I.
(8) Perry argues there is reason to doubt those who claimed that pre-war Canada was a tax haven, citing work by Dr. O.D. Skelton showing Canadas average per capita tax was $31.50, compared with $24.63 in England and $30.90 in the United States (p. 145).
(9) See David B. Perrys Financing the Canadian Federation, 1867-1995: Setting the Stage for Change, Toronto: Canadian Tax Foundation, 1997.
(10) This historical overview is from Gillespie (1991).
(11) The CCH Canadian Ltd. summary of the income tax act including all technical notes, pending amendments, Department of Finance Press Releases, and notices of ways and means weighs in at about 1,890 pages.
(12) Hong Kong residents can choose between paying the flat 15% rate on income or the more traditional tax system, with its attendant deductions and exemptions. Some have proposed this kind of model, called a MAXTAX, for the United States. See Moore (1997) for a discussion of the MAXTAX proposal.
(13) The idea of a flat tax is intricately bound to that of the income tax, even when it exists in opposition to an income tax system (i.e., consumption taxes).
(14) Prior to 1988, there were ten marginal tax rates, ranging from 6% on the first $1,320 to 34% on taxable income exceeding $63,347.
(15) The tax thresholds in 1988 were $27,500 and $55,000 versus $29,590 and $59,180 at the beginning of 2000. That represents a 7.6% increase during a 12-year period compared with a 30% increase in the consumer price index over the same period.
(16) Although tax thresholds (brackets) were indexed to inflation above 3%, this meant little in practical terms because the Bank of Canada pursued an aggressive policy of keeping inflation below this threshold. It was successful for most of the decade.
(17) These were the short-term effects of the tax plan. The medium-term plan was to reduce the middle-income tax rate to 23% and increase tax thresholds to $35,000 for the lowest marginal rate and $35,000 to $70,000 for the middle rate. These higher thresholds are the result of both a basic bump up in the threshold plus full indexation. Indexation also was projected to increase the basic personal exemption to $8,000 from $7,131 within five years. The plan also increased the Canada Child Tax Credit to $2,400 by 2004 and eliminated the 5% surtax for persons with incomes up to $85,000.
(18) The reference to flat tax here means the Hall and Rabushka proposal which appears to be the most thoroughgoing and most discussed of existing proposals.
(19) The rationale is similar to that of taxing dividends only once. Consider, for example, company shares, which in theory represent the capitalized current value of after-tax future earnings. A capital gain is recorded when the price of these shares rise based on expectations of increased future profits. These future profits will be taxed when, and if (markets are not, after all, omniscient), they occur. Because the goal is to tax all income only once, it makes no sense to tax them in the hands of the shareholder who realizes this capital gain.
(20) Recent research in the burgeoning field of economics and psychology suggests that common notions of fairness are severely at odds with the economists vision of fairness.
(21) This income is, however, taxed at the source (i.e., at the corporate level).
Osberg, Poverty in Canada and the USA: Measurement, Trends and
Implications, Presidential address to the Canadian Economics Association, Vancouver,
3 June 2000. Available online at
(23) This is seen as a selling point by some flat tax advocates who want to encourage stay-at-home mothers or fathers.
(24) Inequality may have direct economic impacts as well. For example, J. Barkley Rosser and Marina Vcherashnaya Rosser (2001) present empirical evidence that suggests heightened inequality may encourage the growth of black markets that erode the governments revenue base.
(25) Some economists argue the principle that benefits would come from efficiency gains because the tax effects are nil in the long term.
(26) Note that these are calendar year figures while budgetary projections are generally done in fiscal years ending 31 March. To compensate for this discrepancy, the papers analysis has assumed that one-quarter of the surplus for 1999-2000 and three-quarters of the projected surplus for 2000-2001 represented the equivalent calendar year surplus for 2000. This yielded a projected surplus of $12 billion for calendar year 2000 instead of $11.9 billion for 2000-2001 and $9.2 billion for calendar year 2001 instead of $8.3 billion for 2001-2002.