Great attention has been given recently to changes over time in the average incomes of "quintiles," families or households ranked top to bottom by income and divided into fifths. However, such time line comparisons between rich and poor ignore a central element of the U.S. economy, which is the extent to which individuals move from one quintile to another. Figures on income mobility are more characteristic of the nature of our fluid society than comparisons of average incomes by quintile, which would only be statistically meaningful if America were a caste society where the people comprising the quintiles remained constant over time.
Unfortunately, while data on average income by quintile has been plentiful, however misleading, data on income mobility has been scarce. Until now.
This study is an analysis of newly available panel data based on income tax returns filed from 1979 through 1988, which were tabulated by the U.S. Department of the Treasury. The Treasury sample consists of 14,351 taxpayers filing returns in all of the above years. This sample tends to understate income mobility to the extent the movement of younger and older filers in and out of the population of taxpayers is missed by the requirement that returns be filed in all years. On the other hand, this understatement is at least somewhat offset at the low end of the income scale by the presence of an underclass which does not file tax returns year after year. For the purposes of this report, the bottom quintile consists of those who earn enough income to at least file income tax returns, if not to actually pay taxes.
Earlier studies of income mobility have demonstrated a startling degree of income mobility in as short a period as one year. However, as a January 1992 study noted[1], additional data over more extended periods were needed to draw more precise conclusions about income mobility over the longer term. This need has now been largely satisfied by the provision of longitudinal panel data from tax return files. However, much more data and research on income dynamics in coming years is needed.
The new tax return data support the conclusion of earlier research which concluded that the degree of income mobility in American society renders the comparison of quintile income levels over time virtually meaningless. According to the tax data, 85.8 percent of filers in the bottom quintile in 1979 had exited this quintile by 1988. The corresponding mobility rates were 71 percent for the second lowest quintile, 67 percent for the middle quintile, 62.5 percent for the fourth quintile, and 35.3 percent for the top quintile.
Of those in the much discussed top 1 percent, over half, or 52.7 percent, were gone by 1988. These data understate income mobility in the top 1 percent to the extent mortality contributes to mobility and the diffusion of income. Graph 1 displays the income mobility of the various groups.
In all but the top quintile, at least 60 percent of filers exited their 1979 income quintile by 1988, with two-thirds or more exiting in the bottom three quintiles. Though much more stability was observed in the top fifth, over one-third had slipped downward to be replaced by others moving up. Even most of the top 1 percent had exited by 1988, to be replaced by others.
The very high degree of income mobility displayed above shows that the composition of the various quintiles changes greatly over time. A majority of filers have indeed moved to different quintiles between 1979 and 1988. Thus intertemporal comparisons of average wages, earnings, or private incomes of quintiles cannot provide meaningful measures of changes in the income of actual families and persons only temporarily in a given quintile or percentile. Quintiles may be a convenient way of presenting snapshots of income data for a group of people at a certain point in time. Nonetheless, the notion of a quintile as a fixed economic class or social reality is a statistical mirage.
Movement is important, but the direction of that movement is more important. While a strong argument can be made for a flexible and open market economy which presents opportunities to lower and middle income workers, instability alone is not necessarily a virtue. Graph 2 summarizes the income mobility data to display the direction of movement between 1979 and 1988. For example, in the third, or middle 1979 fifth, 47.3 percent had moved to a higher quintile by 1988, while 33.0 remained in this same quintile, and 19.7 percent fell into a lower quintile.
Given the relative starting position, the very high mobility from the bottom quintile obviously reflects improvement. In addition, the upward movement in the second, third, and fourth quintiles is much larger than downward movement. For example, 60 percent of the second quintile had moved to one of the higher three quintiles by 1988. Over this same time, only 10.9 percent had fallen from the second into the lowest quintile.
In the long overdue debate over the significance of income mobility, some may argue that mobility would tend to reflect slippage, especially among the middle class. The data contradict this contention. Of those in the middle quintile in 1979, nearly half moved upward to the fourth or fifth quintiles by 1988. Overall, in the bottom four quintiles, net improvement was the rule, not the exception.
Table 1 displays the movement of filers from 1979 quintiles to their positions in 1988. Each row can be read across: of 100 percent of each 1979 quintile, the table shows their dispersion among the various fifths by 1988.
About 86 percent of those in the bottom quintile in 1979 had managed to raise their incomes by 1988 enough to have moved up to a higher quintile. The data show that these were not all grouped at the bottom at the second quintile. While 20.7 percent were in the second quintile, 25.0 percent had made it into the middle fifth, and another 25.3 percent into the second highest quintile. The 14.7 percent in the top quintile was actually higher than the 14.2 percent still stuck in the bottom fifth. In other words, a member of the bottom income bracket in 1979 would have a better chance of moving to the top income bracket by 1988 than remaining in the bottom bracket.
In the second quintile, 71 percent had exited between 1979 and 1988. Though 29.0 percent still remained in the second quintile in 1988, 29.6 percent had moved up to the third quintile, 19.5 percent to the fourth, and 11.1 percent to the top quintile. Only 10.9 percent had moved down to the lowest quintile.
Of those in the middle quintile in 1979, 32.3 percent had moved to the fourth quintile and 15.0 percent to the fifth quintile by 1988. Over this period, 47.3 percent had moved up, while 19.7 percent had moved down. The net effect of income mobility in the middle range clearly reflected net overall improvement.
While the fourth quintile exhibited powerful income mobility, the top quintile is the most stable. However, all income mobility from the top quintile is by definition downward mobility. The share of this group dropping into lower quintiles was 35.3 percent, while 27.2 percent of the fourth quintile also dropped at least one quintile. Many of these with declining fortunes are still better off than many of those with upward mobility from a low quintile, however, the overall pattern is that there tends to be strong upward mobility from the lower quintiles, while income mobility from a high level often reflects economic reversals. Without income mobility, many in the top fifth would be better off, and the great majority of those in the lower quintiles would be worse off. Income mobility reflects improvement in the lower four quintiles, but this fact has been virtually ignored in public discussion of income trends.
While 35.3 percent fell from the top quintile into the fourth quintile or below, 40.0 percent of the bottom quintile had moved into the fourth or fifth quintiles by 1988. Of all of those in the bottom quintile in 1979, about two-thirds, or 65 percent, had moved to the middle or higher quintiles by 1988. These data demonstrate that the U.S. economy, not without problems over this period, still remains dynamic, open, and productive enough to permit most Americans in the bottom three-fifths to work their way up the economic ladder. What is needed are policies to ensure that this flexibility and opportunity are extended as widely as possible, especially to those who actually fall below the bottom fifth of taxpayers.
Currently there are two models of the American economy, one static, and the other dynamic. The first portrays the United States as a caste system and misapplies the characteristics of a permanent income strata to those only temporarily moving through income brackets. The alternative view portrays a much more complex and interesting social reality in which the composition of income classes are in constant flux. According to this latter point of view, simplistic generalizations about actual persons and families (or "the rich" and "the poor") cannot be drawn from data on a conceptual artifice which does not exist as such in reality.
The empirical data support the view of the market economy as a dynamic and open society which provides opportunity to those who participate. There is no evidence of stagnation, with the turnover rate in the most stable quintile -- the top fifth -- exceeding 35 percent. The turnover rates in the bottom four quintiles were at least 60 percent over the period, with most of this reflecting upward progress. Analysis which assumes or suggests stable composition of family or household income quintiles rests on invalid assumptions. It makes no sense to draw sweeping conclusions such as "the income of the bottom 20 percent of families fell" in a 15-year period when most of the people originally in that category have long since improved their standard of living enough to have moved up from the bracket entirely.
Christopher Frenze
Senior Economist
Click here to see Graph 1.
Click here to see Graph 2.
Click here to see Table 1.
Return Home