Flat Tax
Principles and Issues

4 Income dynamics and mobility

[Previous] [Contents] [Next]

Tax reforms are most often evaluated in terms of their effects upon particular groups identified by factors such as age, income, or employment status and these effects are most often discussed in the context of a one-year period. Such an analysis leads to erroneous and quite meaningless conclusions and is based on a simplistic view of taxation.

This type of analysis rests on the notion of "fairness," traditionally referred to in public finance as "horizontal and vertical equity." In fact, the question of fairness, however defined, is often seen as a prerequisite to assessing the appropriateness of a particular reform to the tax system. Critics of the reform point to the "unfairness" of the change to an identifiable group. This unfairness then becomes an insurmountable barrier that, no matter how large the benefits of the change, no amount of evidence can dislodge.

Argumentation based on unfairness implicitly accepts the status quo as a reasonable representation of a fair (or fairer) state. Those who accept this proposition often incorrectly use the impact on the tax system in a single year as the basis for analysis, when a horizon encompassing several years is more appropriate. When looking at a single year of data, a casual examination of the impact of a revenue-neutral flat-tax system or the replacement of multiple rates of tax with a single rate leads to one obvious conclusion: such a system leads to lower taxes for those earning a high income and higher taxes for those with low and middle incomes.

To see such an analysis is premature, we need only to look at the importance of income mobility. It is obvious to most people that at different times of our lives we have different incomes. Those first entering the work force or enrolled in educational programs (typically the young), understandably have lower incomes. In fact, most of the authors of this Critical Issues Bulletin lived below Statistic Canada's Low Income Cut Off line during their time in university. Similarly, retired persons drawing down retirement savings would also be classified as those with low income. Persons in this age category typically have paid off their debts and thus require less income for the same standard of living as they had during their working life. Between these two extremes lies a pattern of first increasing, then decreasing income (Gunderson and Riddell 1993: 175). The movement of people from one category of age and income to another makes it difficult to defend statements such as "the condition of those in low-income groups has deteriorated over the past ten years" because people with a low income in year one do not, for the most part, have a low income in later years.¹⁷ It is similarly difficult to make statements like "the poor are going to bear the burden of this tax reform" because those who are poor today are likely to be relatively well off tomorrow.

Nature of income mobility

Statistics Canada's Survey of Labour and Income Dynamics (Webber, Cotton, Meere, Bishop, and Hewer 1999) can help us understand the movement between income groups better. This survey tracks specific persons over time and allows us to observe and to monitor how a person's income changes over time. This type of longitudinal analysis is often carried out by grouping people into income quintiles¹⁸ according to their annual earnings. The study then examines how the composition of each income group (quintile) changes over time.

Canadian Results

Webber et al. (1999) examined Canadian panel data to investigate the extent of income mobility. The following summarizes their findings on income mobility for a two-year period:

• 66.5% of families did not change quintile
• 13.8% moved up one quintile
• 13.1% dropped one quintile
• 3.2% moved up more than one quintile
• 3.5% dropped more than one quintile.

Of those families initially in the bottom two quintiles in 1995, 24% found themselves at least one quintile higher by 1996. In other words, nearly one-quarter of families in the lowest two income groups had moved up at least one income group within a one-year period. This depicts Canadian income mobility as fairly dynamic. In fact, after examining the available Canadian data, Emes and Walker (1999) conclude that "there is not a `permanent underclass' stuck in the lower income group" (1999: 54).

When a similar analysis is extended to a five-year period (Webber, Cotton, Meere, Bishop, and Hewer 1999), the results show, as expected, even greater income mobility. The data shows that:

• 49.1% of families did not change quintile
• 20.7% moved up one quintile
• 14.5% dropped one quintile
• 8.2% moved up more than one quintile
• 7.5% dropped more than one quintile

Of those initially in the bottom two quintiles, 45% moved up at least one quintile over the five-year period of the study. In other words, in only 5 years nearly half of all Canadian families in the bottom 40% moved up at least one income group. In fact, almost as many families moved up a quintile as stayed in the same quintile. Should the data be extended further, even greater levels of income mobility can be expected.¹⁹

Research into British income mobility corroborates Canadian results. The research examined the early 1990s and found "much mobility in household net income from one year to the next" (Jarvis and Jenkins 1998: 428). They also found relatively greater mobility in the low and high ends of the income distribution than in the middle. This is, again, suggestive of a dynamic income distribution.

Research in the United States

Unfortunately, both the Canadian and British data sets are limited in their applicability to the current discussion because they cover very few years. A more accurate sense of income mobility can be found in the University of Michigan's Panel Survey on Income Dynamics, the results of which were presented in a recent report by the Federal Reserve Bank of Dallas (1995). The study, which looked at incomes during the period 1975 to 1991, found the following key facts.

• Only one-half of one percent (0.05%) of the sample was in the bottom quintile for every year of the study, suggesting that "being in the low-income bracket isn't, for a large majority of people, permanent" (1995: 8).
• Only 5.1% of those in the bottom quintile in 1975 were in the bottom quintile in 1991.
• The average income in the poorest quintile grew 2,196% from 1975 to 1991 while the average income in the richest quintile grew 8.7% in this time period.
• The young and the educated were more likely to move up the income distribution.
• The report concludes: "All through the University of Michigan data, there's a consistent, powerful thrust toward the top of the income distribution" (1995: 8).

The report cites a study from the US Treasury Department examining tax returns from over 14,000 people from 1979 to 1998. The study confirms many of the findings of the University of Michigan. For instance, "86% of those in the lowest income bracket moved to a higher grouping. Two-thirds of them reached the middle strata or above, with almost 15% making it all the way to the top fifth of income earners" (Federal Reserve Bank of Dallas 1995: 12). Interestingly, of those in the low income group in 1979, more had made it all the way to the top of the income distribution in those nine years than remained in the bottom quintile.

Census Bureau data for the 1980s consistently shows roughly 20% of the people in the bottom quintile moving up a minimum of one quintile within a one-year period. Interestingly, the data also showed roughly the same percentage of people in the top quintile of earnings moving down a minimum of one quintile (cited in Bartlett 2000).

Finally, a study from the Urban Institute (Steuerle 1998) concluded that approximately one-half of those in the bottom quintile during the period from 1967 to 1976 had moved up when surveyed in the period from 1977 to 1986. Similar to the findings of the Census Bureau, an almost equal number of individuals in the top earnings quintile had moved down over the same two periods. The results of these studies speak volumes for the importance of considering income mobility when analyzing tax reforms.

Implications for the flat tax

Even when evaluating tax reform against the status quo, static analysis is only the starting point. A blanket statement that this reform helps the rich or hurts the poor cannot be justified. One year's data provides insufficient information to conclude that there is systematic "unfairness" given the movement of taxpayers from one income group to another. A rational taxpayer would be concerned about his or her lifetime tax burden, not how a particular change will affect next year's tax bill. Most people expect to earn a greater income in their 40s and 50s than they do in their 20s and 30s; most do so. A rational taxpayer would be willing to trade-off a slightly larger tax burden today for the efficiency gains, increased economic growth, and much larger tax savings in the future based on tax reform. This is no different from saving money today to spend tomorrow. The effects of a change in the tax regime, therefore, should be looked at over a number of years.

Life-cycle analysis

A life-cycle analysis by Walker (1983) of various flat-rate tax proposals provides us with a foundation for analysis. Walker notes that, of those in the main group of income earners who would be paying greater taxes, a full 63% would be in periods of their lives with abnormally low levels of income (due to youth or old age) relative to the tax system that existed at the time of the analysis. For the portion of those income earners who were young, their tax burden would decline as they aged, offsetting the effects of an initial increase. Walker also finds that the burden of lifetime taxes for all but one of the flat-rate tax systems he examined is equivalent to, or less than, the burden of lifetime taxes from the tax system in existence at the time of analysis.

Tables 4 and 5 show the changes in income tax for each of the cases analyzed in Section 3 compared with the current tax system. All of the cases except cases 8 and 9, which include spending and tax reductions, show an increased tax burden during the early years of the taxpayers working life as well as during the later years of life. This increased burden is offset by a decreased burden during mid-life, usually the years from age 30 to age 64.²⁰ Let us examine the experience of an average individual ²¹ in Case 3, which offers a personal exemption of $8,766 with no tax credits or other deductions. Before turning 20, the average individual is burdened with additional income taxes of $12 per year above the status quo. This burden increases with age, peaking at an extra $214 during the years from age 25 to age 29. It then drops to $68 above the status quo during the years from age 30 to age 34. This is followed by many years (from 35 to 64) when the tax bill is less than it would have been under the current system. In some cases, these reductions are substantial. For instance, during the years from age 55 to age 59, income taxes are over $800 less in Case 3 than under the status quo. Beyond 64 years of age, the average individual pays over $200 more in tax per year.

This example illustrates several important points. First, there is a relatively small amount of extra income tax paid during those years in which taxes paid in Case 3 are higher than the status quo. In fact, before age 30 the average increase in taxes amounts to around 0.8% of income. After 65, the increase in taxes is 1.3% of total income. Second, the period during which taxes are lower is longer than the periods when taxes are higher and the reductions in taxes are typically larger than the increases. During the period from 35 years to 65 years of age, the reduction in taxes is, on average, 1.3% of income. Over this individual's lifetime, the biggest tax decrease is 2.4% of income, whereas the biggest increase is 1.6%. These numbers indicate that the transition to a flat tax like that analyzed in Case 3 would be, for the average individual, relatively painless and most likely profitable.

The bottom line for an average individual in Case 3 is a lifetime reduction in income taxes of $6,408 (table 6). As shown in table 6, all of the Cases presented in Section 3 result in a lifetime reduction in income-tax liability for an average income earner. This is obviously due to the fact that the reductions obtained during the person's peak earning period more than offset the tax increases incurred during periods of lower income.

If the status quo is the ultimate arbitrator of fairness, the evidence presented suggests that the flat tax as analyzed in Cases 1 through 9 constitute a fairer tax system. Those with low incomes may temporarily bear the burden of the change in tax systems but when they begin to earn higher income, as research shows they almost certainly will, they will enjoy lower rates of income tax that will more than compensate them for the relatively small increases incurred in their younger years.²² Other sections of this study highlight the important benefits gained through tax reform based on a flat tax and, to a lesser extent, on a single-rate tax. This analysis shows that the lifetime tax benefits do not come at the cost of fairness to those at various points in the income distribution.

Table 4: Income tax payable--status quo versus single-rate tax reform (Cases 1 through 9)

Table 5: Income tax payable--change from status quo (Cases 1 through 9)

Other considerations

Another consideration is the impact that changing the tax system will have on income mobility. If the changes induce greater work effort or greater investments in education,²³ the average taxpayer will be much better off under the reformed system. Is there evidence that this would happen?

Samida (1998) examines the relationship between economic freedom in Canada and the percentage of people in a particular age group with incomes under $12,500 (in real terms). He finds that the greater the level of economic freedom a province has, the greater the percentage of people who will surpass this level of income seven years later. In so much as the single-rate tax or a flat tax reduces marginal tax rates and, by implication, increases the level of economic freedom, we might expect income mobility to increase as well.

Table 6: Change in income tax liability over lifetime (Cases 1 through 9)

Conclusion

Given that income increases and decreases over an individual's lifetime, it is clear that an entire lifetime is the proper framework in which to analyze tax reform. We should assess the effect of a change in the tax system upon a person's lifetime tax liability rather than simplistically and quite incorrectly assessing its effect in a particular year. Further, the data available for Canada, Britain and, in particular, the United States strongly suggest that, in all three jurisdictions, individuals move from one level of income to another to a relatively high degree.

[Previous] [Contents] [Next]

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.