Many people are aware that income inequality has increased over the past twenty years. Upper-income groups have continued to do well, decade after decade, while others, particularly those without a college degree, and especially the young, have seen their real income decline. The 1994 Economic Report of the President refers to the 1979-90 ' fall in real income of men with only four years of high school-a 21 percent decline-as "stunning."' Indeed, these changes in income distribution constitute one of the strongest arguments for the tax measures in President Clinton's 1993 deficit reduction program. Those at the top of the income distribution who are being asked to contribute more are the very people whose incomes grew most rapidly.
This paper broadens the discussion of distribution and taxation to include something that usually is neglected: wealth. Almost all discussions of distributional issues have centered on income. Income in any year is a measure of a household's economic. position but it can vary greatly from year, to year. A high income still can leave a household vulnerable. A better indicator of long-run economic security is wealth, the net worth of the household. Wealth is found by adding together the current value of all the assets a household owns-financial wealth such as bank accounts, stocks,. bonds, life insurance savings, mutual fund shares; houses and unincorporated businesses; consumer durables like cars and major appliances; and the value of pension rights-and subtracting liabilities=consumer debt, mortgage balances, other out standing debt. Wealth can vary from year to year as asset prices rise and fall, but it remains the foundation for a family's long-term security. Without wealth, a family lives from hand to mouth, no matter how high its income.
Examination of the data on wealth distribution leads to a disturbing question: Is America still the land of opportunity? The growing divergence evident in income distribution is even starker in wealth distribution. Equalizing trends of the 1930s-1970s reversed sharply in the 1980s. The gap between haves and havenots is greater now than at any time since 1929. The sharp increase in inequality since the late 1970s has made wealth distribution in the United States more unequal than in what used to be perceived as the class-ridden societies of northwestern Europe.
Contrary to popular perception, -the go-go years of the 1980s did not offer everyone a piece of the action. They were a party for those at the very top of the wealth distribution. While those in the fast lane enjoyed large increases in wealth during the 1980s, the wealth of the rest of the population did not simply grow more slowly; it actually fell. Looking at real financial wealth alone-including bank accounts, stocks, and bonds but excluding durable goods, housing, and pension wealth-80 percent of households experienced a decline between 1983 and 1989. Adding in housing wealth makes the picture somewhat less grim, but the bottom 40 percent of households still had less wealth in 1989 than in 1983.
Not surprisingly, in light of these facts, the racial distribution of wealth deteriorated in the 1980s from an already unacceptable level. Relative income of African American households held steady at about 60 percent of white income in the 1980s, but the relative wealth position of most black families deteriorated. Historically, black wealth always has been much lower than that of whites, the legacy of slavery, discrimination, and low incomes. Between 1983 and 1989, a bad situation grew worse. In 1983, the median white family had eleven times the wealth of the median nonwhite family. By 1989 this ratio had grown to twenty. Middle-class black households did succeed in narrowing the wealth gap with whites, but most nonwhite families moved even further behind. More than one in three nonwhite households now have no positive wealth at all, in contrast to one in eight white households.
These trends suggest that for reasons of fairness, the United States should consider broad taxation of wealth. Currently, we tax wealth in several specific ways. Our most important wealth tax, the property tax, is administered at the state and local level and is beyond the scope of this study. At the federal level, wealth. is taxed through the capital gains tax and, at death, through estate taxes. Eleven other OECD countries have additional taxes on wealth. By studying these taxes and employing computer simulation, it is possible to assess the revenue effects and distributional implications that would be obtained were the tax systems of Germany, Sweden, or Switzerland to be adopted in the United States.
Although wealth taxation in one form or another is ubiquitous in rich countries, nowhere is it a major revenue source (that is, more than a small percentage of total revenue). The international mobility of financial wealth and widespread concern about the incentive effects of wealth taxation-incentives against saving and for capital flight-as well as the power of affluent elites all work to reduce the level of effective taxation. Nevertheless, even a very simple and modest tax- like Switzerland's could raise substantial revenues in the United States. A wealth tax modeled on Switzerland's-with a $100,000 exclusion and a top rate for the wealthiest of three-tenths of one percent-would generate an estimated $40 billion annually. This is not much when measured against total federal revenues, to say nothing of total income or wealth. Nevertheless, it is significant in the context of the debates surrounding the Clinton budgets. A wealth tax offers us an important new fiscal option. It could be used to help finance any of the many needs of our society, from highway and bridge repair to environmental cleanup to health care to prisons. Or it could be used to replace other taxes that are deemed less fair or entail more damaging economic disincentives. And while any new tax would bring howls from the antitax lobby, given the concentration of financial wealth, it would be hard to find a lot of people who would be directly affected. Even the wealthiest households could be expected to pay less than 10 percent of the annual real return on their capital as their new wealth tax liability; more than three-quarters of all families would pay practically nothing. Would such a tax have terrible supply-side effects? Well, one might ask whether Switzerland suffers from capital flight and impoverishment.
2. WHY WEALTH?
Before exploring more closely the facts and policies surrounding wealth inequality, a review of the central concepts is helpful. Family wealth refers to the net dollar value of the stock of assets less liabilities (or debt) held by a household at one point in time. Income, in contrast, refers to a flow of dollars over a period of time, usually a year. Though certain forms of income are derived from wealth, such as interest from savings accounts and bonds, dividends from stock shares, and rent from real estate, income and wealth are by no means identical. Many kinds of income--wages and salaries, food stamps and other transfer payments--are not derived from household wealth, and many forms of wealth, such as owner-occupied housing, produce no corresponding cash income flow.
Most people think of family income as a measure of well-being, but family wealth is also a source of well-being, independent of the direct income it provides. There are both narrowly economic and broader reasons for the importance of wealth. Some assets, particularly owner-occur pied housing, provide services directly to their owner. This is also true for consumer durables, such as automobiles. Such assets can substitute for financial income in satisfying economic needs. Families receiving the same financial income but differing in their stocks of housing and consumer durables will experience different levels of well-being.
More important, perhaps, than its role as a source of income is the security that wealth brings to its owners, who know that their consumption can be sustained even if income fluctuates. Most assets can be sold for cash or used as collateral for loans, thus providing for unanticipated consumption needs. In times of economic stress, occasioned by such crises as unemployment, sickness, or family breakup, wealth is an important cushion. The very knowledge that wealth is at hand is a source of comfort for many families.
In the political arena, large fortunes can be a source of economic power and social influence that is not directly captured in the measure of annual income. Large accumulations of financial and business assets can confer special privileges on their holders. Such fortunes are often transmitted to succeeding generations, thus creating family "dynasties."
In most households, wealth varies from year to year. First, saving out of current income may augment wealth, just as spending in excess of income may diminish it. Second, assets already held by the family may change in value! Finally, gifts and inheritances from or to a family member may change household wealth.
It is apparent that a family's wealth can be expected to depend on the age of its members because older working individuals generally will have spent more years saving and accumulating assets. Indeed, wealth, like income, is related to age, but the relationship is not a strong one since savings rates, the rate of return on asset holdings, and gifts and inheritances will generally differ among families of similar age profile, even when they have the same earnings history.
Wealth and income are positively correlated (that is, families with more income generally have more wealth), but this association, too, is far from perfect. One reason is that rates of return on various components of wealth vary widely between years and within any one year.' Even wellreported asset and income data would yield an incomplete picture of the wealth from which they flow. Other types of wealth, as noted above, may yield no income at all in a given year. Age explains part of the variation of the wealth-income ratio, but much is left unaccounted for. As a result of this unexplained variability, wealth measures well-being differently from annual income, both in relative and absolute terms.
3. HOUSEHOLD WEALTH INEQUALITY IN THE UNITED STATES: PRESENT LEVEL AND HISTORICAL TRENDS
Wealth inequality in the United States was at a sixty-year high in 1989 (the latest date available), with the top 1 percent of wealth holders controlling 39 percent of total household wealth. Focusing more narrowly on financial wealth, the richest 1 percent of households owned 48 percent of the total. How did this come to pass? After the stock market crash of 1929, there ensued a gradual if somewhat erratic reduction in wealth inequality, which seems to have lasted until the late 1970'. Since then, inequality of wealth holdihgs, like that of income, has risen sharply (see Figure 3-1).' If Social Security and other pension wealth are included ("augmented wealth"), the improvement between 1929 and 1979 appears greater, but the increase in inequality since 1980 is still sharply in evidence.
The rise in wealth inequality from 1983 to 1989 (a period for which there is comparable detailed household survey information) is particularly striking. The share of the top 1 percent of wealth holders rose by 5 percentage points. The wealth of the bottom 40 percent showed an absolute decline. Almost all the absolute gains in real wealth accrued to the top 20 percent of wealth holders.
CHANGES IN AVERAGE WEALTH HOLDINGS
Look first at trends in real wealth over the period from 1962 to 1989 (all in 1989 dollars). As Figure 3-2 shows, average wealth grew at a respectable pace from 1962 to 1983 and even faster thereafter. In fact, mean marketable wealth grew almost twice as fast between 1983 and 1989 as between 1962 and 1983 (3.4 percent per year vs. 1.8 percent). By 1989, the average wealth of households was $200,000, almost double that of 1962.
Average financial wealth grew faster than marketable wealth in the 1983-89 period (4.3 versus 3.4, percent per year), reflecting the increased importance of bank deposits, financial assets, and equities in the overall household portfolio over this period. This reversed the relationship of the 1962-83 period, when financial wealth grew more slowly than marketable wealth (1.4 versus 1.8 percent per year).
Average household income also grew faster in the 1983-89 period than in the 1962-83 period. Its annual growth accelerated from 1.5 percentage points to 2.7. However, in both periods average income grew more slowly than average wealth, with the difference rising from 0.3 percentage points per year for 1962-83 to 0.8 for 1983-89.
The robust growth of average wealth disguises some changes in the distribution of that wealth. This becomes clear after examination of median rather than mean wealth. Mean wealth is simply the average: total wealth divided by total number of households. If the wealth of only the top 20 percent of households increases (with nothing else changing), then mean wealth increases because total wealth increases.' In contrast, the median of the wealth distribution is defined as the level of wealth that divides the population of households into two equal-sized groups (those with more wealth than the median and those with less). Returning to the earlier example, if only the top quintile enjoys an increase in wealth, median wealth is unaffected even though mean wealth increases because all additional wealth accrues to people well above the median income. The median tracks what is happening in the middle of the wealth distribution.' When trends in the mean deviate from trends in the median, this is a signal that gains and losses are unevenly distributed.
The trend in median household wealth in the United States gives a contrasting picture to the growth of mean wealth. Unlike mean marketable wealth, median marketable wealth grew faster in the 1962-83 period than in the later period (1.6 percent per year vs. 1.3 percent in 1983-89). Median wealth also grew much more slowly than mean wealth in the later period (a difference of 2.2 percentage points per year). Overall, from 1983 to 1989, while mean wealth increased by 23 percent, median wealth grew by only 8 percent. The fact that mean wealth grew much faster than median wealth after 1983 implies that the bulk of the gains were concentrated at the top of the distribution-a finding that implies rising wealth inequality.
RISING WEALTH INEQUALITY IN THE 1980s
The rising level of wealth inequality between 1983 and 1989 is illustrated in Figure 3-3. The most telling finding is that the share of marketable net worth held by the top 1 percent, which had fallen by ten percentage points between 1945 and 1976, rose to 39 percent in 1989, compared with 34 percent in 1983.6 Meanwhile, the share of wealth held by the bottom 80 percent fell by more than a fifth, from 19 to 15 percent.
These trends are mirrored in financial net worth, which is distributed even more unequally than total household wealth. In 1989, the top 1 percent of families as ranked by financial wealth owned 48 percent of the total (in contrast to 39 percent of total net worth). The top quintile accounted for 94 percent of total financial wealth, and the second quintile accounted for nearly all the remainder.
The concentration of financial wealth increased to the same degree as that of marketable wealth between 1983 and 19$9. The share of the top 1 percent of financial wealth holders increased by 5 percentage points, from 43 to 48 percent of total financial wealth. The share of the next 19 percent fell from 48 to 46 percent, while that of the bottom 80 percent declined from 9 to 6 percent.
Income distribution, too, became more concentrated between 1983 and 1989. As with wealth, most of the relative income gain accrued to the top 1 percent of recipients, whose share of total household income grew by 4 percentage points, from 13 to 17 percent. The share of the next 19 percent remained unchanged at 39 percent. Almost all the (relative) loss in income was sustained by the bottom 80 percent of the income distribution, whose share fell from 48 to 45 percent.
Another way to view rising wealth concentration is to look at how the increases in total wealth were divided over a specified period. This is calculated by dividing the increase in wealth of each group by the total increase in household wealth.' The results, shown in Figure 3-4, indicate that the top I percent of wealth holders received 62 percent of the total gain. in marketable wealth over the period from 1983 to 1989. The next 19 percent received 37 percent, while the bottom 80 percent received only 1 percent. This pattern represents a distinct turnaround from the 196283 period, when every group enjoyed some share of the overall wealth growth and the gains were roughly in proportion to the share of wealth held by each in 1962. Over this period, the top 1 per- ' cent received 34 percent of the wealth gains, the next 19 percent claimed 48 percent, and the bottom 80 percent got 18 percent.
Gains in the overall growth in financial wealth were distributed even more unevenly than in marketable net worth, with 66 percent of the growth accruing to the top 1 percent and 37 percent to the next 19 percent. The bottom 80 percent collectively lost 3 percent.
Finally, the changes in wealth distribution can be assessed by looking at the Gini coefficient. This indicator is used commonly to summarize data on the degree of inequality of income; wealth, or anything else of value. It ranges from 0 (exact equality) to 1 (one person owns everything); a higher Gini coefficient means greater inequality. This measure, like the others reviewed here, points to an increase in inequality: between 1983 and 1989 the Gini coefficient increased from 0.80 to 0.84.
The increase in wealth inequality recorded over the 1983-89 period in the United States is almost unprecedented. The only other period in the twentieth century during which concentration of house- E , hold wealth rose comparably was from 1922 to 1929. Then inequality was c.Abuoyed primarily by the excessive increase in stock values, which eventually crashed in 1929, leading to the Great Depression of the 1930s.
4. THE CHANGING STRUCTURE OF HOUSEHOLD WEALTH
In order to assess the likely incidence of wealth taxation in the United States-who is likely to pay and how much-it is necessary to analyze the variation of household wealth by demographic group and the composition of wealth. There are substantial differences in wealth holding by demographic category. Households in the age group 45-69' are by far the wealthiest group in our country, with those 70 and over in second place and households under 45 a distant third. Between 1983 and 1989 the two less privileged age cohorts made gains in relative wealth holdings. Over the same period, white and nonwhite households' mean wealth continued to converge. Nevertheless, the gap in mean wealth holdings between whites and nonwhites remained very large in comparison with income differences (a ratio of 0.29 for wealth versus 0.63 for income). Moreover, the lower half of the distribution for nonwhite households actually fell further behind the lower half for white households (median wealth diverged), hinting that wealth inequality among nonwhites has been growing even faster than among the population at large.
WEALTH HOLDINGS BY AGE GROUP
One key predictor for differences in wealth among families is age. Individuals typically accumulate wealth until retirement age; thereafter, they spend down their savings. Figure 4-1 confirms this pattern. In 1983, the average marketable wealth of families under the age of forty-five was far below the overall average, while that of families over forty-five was well above the average, though less so in the case of elderly households. "Hump-shaped" profiles are also found for 1962 and 1989.
Despite the overall similarity in the three age-wealth profiles, there have been notable shifts in relative wealth holdings. Between 1962 and 1983, middle-aged households gained at the expense of both the younger and older households. The wealth of families under forty-five declined from 50 percent of average to 40 percent, that of the middleaged increased from 135 to 173 percent of average, and that of the older group fell from 146 to 121 percent of the average.
The years 1983 to 1989 saw an exact reversal of this pattern. The wealth of families under forty-five increased from 40 percent of the overall mean in 1983 to 48 percent in 1989, close to the corresponding 1962 figure. The relative wealth of the middle-aged fell from 173 percent of average in 1983 to 159 in 1989, still above the comparable 1962 figure. The wealth holdings of older households grew from 121 percent in 1983 to 139 percent in 1989, remaining below their corresponding level in 1962.
RACIAL DIFFERENCES IN HOUSEHOLD WEALTH
On the basis of the Current Population Reports, the relative gap in income between nonwhite and white households was almost identical in 1967 and 1989 (see Table 4-1). The ratio of mean household income remained at 0.63, and the ratio of median income held at slightly below 0.6. The story of household wealth is more discouraging. From 1983 to 1989 the ratio of mean wealth increased (from 0.24 to 0.29), but the ratio of median wealth fell from the already shockingly low level of 0.09 to only 0.05, leading to the inference that most of the gains for nonwhite households were captured by the upper half of the distribution rather than those less well off.
Notably, the home ownership rate based on decennial census data almost doubled among nonwhite families between 1940 and 1980 (from 24 percent to 44 percent). Indeed, the ratio of home ownership rates between nonwhites and whites increased from 52 percent in 1940 to 64 percent in 1985. However, these increases were confined to the 1940s and the 1960s. Since 1970, there has been no increase in the home ownership rate among nonwhite families. According to the Survey of Consumer Finances (SCF) data, it actually fell slightly between 1983 and 1989, from 43.6 to 43.3 percent.
Turning to wealth more broadly defined, one finds that nonwhite families also made substantial gains on whites in terms of both mean and median net worth between the early 1960s and the early 1980s. According to the SCF data, the gap in mean wealth closed further during the 1980s. Yet the gap in median wealth widened. This result reflects a greater inequality in wealth among nonwhites than whites. In 1989, for example, 35 percent of nonwhite families reported zero or negative net worth, compared to 12 percent of whites. Thus, though there have been some gains in closing the racial wealth gap among better-off nonwhites, the differential is large and growing for the median family.
THE COMPOSITION OF WEALTH
The portfolio composition of household wealth shows the forms in which households save. Do households save for direct consumption, as in acquiring ownership of houses and automobiles? Do families save for precautionary reasons, as in the form of bank deposits? Do they save for retirement, as in insurance plans, IRAs (individual retirement accounts), or the like? Or do they save mainly for investment purposes, as in financial securities and corporate stock?
Overall, between 1962 and 1989 there was a major shift in household portfolios out of financial assets and equities (deposits, bonds, stocks, and trusts), which declined from 52 percent of gross wealth to 38 percent, and a corresponding increase in real estate and unincorporated business equity, rising from 48 percent to 59 percent. Debt as a proportion of net worth, after falling from 16.4 to 15.1 percent between 1962 and 1983, increased to 16.5 percent in 1989.
Many people believe that housing (more specifically, owner-occupied housing) is by far the most important asset the household controls. Owneroccupied housing was indeed the most important asset in the household portfolio in 1962, 1983, and 1989 (see Figure 4-2). However, in none of the three years was its gross value more than a third of total assets, or its net value more than one-quarter. In 1989, housing accounted for 29 percent of the gross value of assets, and net equity in owner-occupied housing-the value of the house minus any outstanding mortgage-was only 20 percent of gross assets (or 24.9 percent of net worth). Checking deposits, savings accounts (including money market funds), and other deposits (including retirement plans like IRAs)' amounted to 17 percent. Real estate other than owneroccupied housing and unincorporated business equity comprised 29 percent of total assets. Corporate stock, bonds and other financial securities, and trust equity amounted to 21 percent. Debt as a proportion of gross assets was 14 percent.
There have been some significant changes in the composition of household wealth since 1962. Popular perception is that housing is the only substantial asset that most families can claim, but figures show this is increasingly untrue. The gross value of housing as a proportion of gross assets increased from 26 percent in 1962 to 30 percent in 1983 but then declined slightly to 29 percent in 1989. Other (nonhome) real estate and business equity grew rapidly between 1962 and 1989, from 22 to 29 percent of total assets. Liquid assets, including checking and savings accounts, money market funds, CDs, life insurance, and pension accounts, remained relatively steady at 20 and 19 percent of total assets in 1962 and 1983, respectively, falling to 17 percent in 1989. Financial securities (bonds), corporate stock, and trust equity declined in importance in the household portfolio from 32 percent in 1962 to 16 percent in 1983, and then increased during the latter part of the 1980s to 21 percent in 1989.
5. COMPARISONS WITH OTHER COUNTRIES
Due to differences in reference sources, definitions of household , ., wealth, and accounting conventions, international comparisons of household wealth inequality must be made cautiously. However, the G.= evidence seems to suggest that in the early part of this century, (the h ° 1920s are the earliest period for which data are available), wealth h inequality was much lower in the United States than in the United Kingdom, while U.S. figures were comparable to Sweden. America appeared to be the land of opportunity, whereas Europe was a place where an entrenched upper class controlled the bulk of wealth. By the late 1980s, the situation appears to have completely reversed, with much higher concentration of wealth in the United States than in Europe. Europe now appears the land of equality.
UNITED KINGDOM AND SWEDEN. There are two other countries besides the United States for which long-term time-series are available on household wealth inequality: the United Kingdom and Sweden. The most comprehensive data exist for the United Kingdom. The data are based on estate duty (tax) return data and, as a result, use mortality multipliers to obtain estimates of the wealth of the living (see Appendix for methodology). Estimates are for the adult population (that is, individuals, not households). Figures are available on an almost continuous basis from 1923 to 1990.
The Swedish data are available on a rather intermittent basis from 1920 through 1990. The data are based on actual wealth and tax returns. Tax return data are saubject to error, like other sources of wealth data. The principle problem is underreporting owing to tax evasion and legal tax exemptions. However, some assets, such as housing and stock shares, are extremely well covered because of legal registration requirements in Sweden. Also, the deductability of interest payments from taxable income makes it likely that the debt information is very reliable. On the other hand, bank accounts and bonds are not subject to similar tax controls, an it is likely that their amounts are underreported.
Figure 5-1 shows comparative trends among the three countries. For the United Kingdom, there was a dramatic decline in the degree of individual wealth inequality from 1923 to 1974 but little change thereafter. Based on a conventional definition of wealth (marketable wealth), the share of the top 1 percent of wealth holders fell from 59 percent in 1923 to 20 percent in 1974. However, between 1974 and 1990, there was only a relatively minor reduction in the concentration of household wealth, as the top percentile saw their share shrink to 18 percent.
In Sweden, as in the United Kingdom, there was a dramatic reduction in wealth inequality between 1920 and the mid-1970s. Based on the years for which data are available, the decline appears to have been a continuous process. Over this period, the share of the top percentile declined from 40 percent to 17 percent of the total household marketable wealth. Between 1975 and 1985, there was virtually no change in the concentration of wealth. However, this was followed by a sharp increase in wealth inequality between 1985 and 1990, with the share of the top percentile increasing to 21 percent, a level similar to that of the early 1960s.
Comparisons among the three countries are instructive. In all three countires, there was a fairly sizable reduction in wealth concentration until the late 1970s, though the pattern was much more cyclical in the United States than in the other two. However, duing the 1980s, both the United States and Sweden showed a rather sharp increase in wealth inequality, whereas the trend was almost flat in the United Kingdom. This is a surprising difference, since both the United States (under Ronald Reagan) and the United Kingdom (under Margaret Thatcher) pursued conservative economic policies while the Social Democrats dominated Sweden. Moreover, of the three countries, Sweden is the only one with a direct tax on household wealth. This suggests that differences in public policy alone cannot account for these trends in wealth distribution.
6. COMPARISONS WITH INCOME INEQUALITY
Wealth inequality is today and has always been extreme and substantially greater than income inequality. Indeed, the top 1 percent of wealth holders has typically held in excess of one-quarter of total household wealth, in comparison to the 8 or 9 percent share of income received by the top percentile of the income distribution. Figure CU 6-1 shows the historical pattern of wealth and income inequality through means of the shares held by the most prosperous families.
Thirty-seven percent of the total real income gain between 1983 and 1989 accrued to the top 1 percent of income recipients (in contrast to 62 percent of the marketable wealth gain), 39 percent went to the next 19 percent of the income distribution, and 24 percent accrued to the bottom 80 percent (versus only 1 percent of the marketable wealth gain). While not as powerfully as in the case of wealth, these results for -income show again that the growth in the economy during the 1980s was concentrated in a surprisingly small part of the population. To put it succinctly, the top quintile received more than three-quarters of the total increase in income and essentially all of the increase in wealth. The starkness of these numbers suggests a widening fissure seaprating the strata within our society.
Though wealth is more unequally distributed, the historical course of wealth distribution has roughly paralleled that of income distribution. Income inequality, as measured by the share flowing to the top 5 percent, increased between 1922 and 1929-from 21.2 to 25.1 percent-declined steadily during the depression years-reaching a 21.8 percent share in 1939-and then fell precipitously during World War II. There was a slight decline between 1945 and 1953, but then income inequality remained virtually flat until 1981. Between 1981 and 1989, it rose fairly sharply, from 15.4 to 17.9 percent.
The comparison of trends in income inequality and wealth inequality during the 1980s is revealing, particularly since the former has received so much attention in both professional economic journals and the mass media. The evidence presented above indicates that the level of wealth concentration was at a postwar high in 1989. The time series on income inequality indicates exactly the same result for the concentration of household income. Moreover, the run-up in wealth inequality that characterized the 1980s had a twentieth-century precedent only during the 1920s. A similar finding can be reported for income inequality.
It is hard to provide a direct comparison of the degree of increase in inequality for the two series because of limitations on data availability. The Gini coefficient for wealth inequality shows an increase from 0.80 in 1983 to 0.84 in 1989. On the basis of the Current Population Report series, the coefficient for income inequality rose from 0-41 to 0.43 over the same period.' The share of wealth held by the top 5 percent of wealth holders increased from 56 to 61 percent over these-years, whereas the share of total income received by the top 5 percent of income recipients moved upward from 17.1 to 18.9 percent. The change in wealth inequality was more pronounced even if the years in question are 1977 to 1989, which includes the entire period of growing income inequality. Over this period, the Gini coefficient for income inequality increased by only 0.03, and the income share of the top 5 percent by only 2.1 percentage points.
It was reported above that the bottom two quintiles of wealth holders experienced an absolute decline in average net worth (in real terms) between 1983 and 1989. Were trends in real income comparable? According to the Current Population Report series, the mean income of each of the bottom two quintiles increased in real temps over the period (by 11 percent and 10 percent, compared to 27 percent for the top 5 percent of the distribution). However, it is striking that the average real incomes of both the bottom quintile and the second-lowest were actually slightly less in 1989 than in 1973, even though overall mean income for the entire population had grown over this period.' Thus, poorer households have seen their net worth declining at the same time as their incomes were stagnating.
It is apparent that the time patterns of wealth inequality and income inequality have been similar but not identical across the century. Income inequality was generally stable during the postwar yearsuntil 1981, while wealth inequality fell sharply during the 1970s. What can explain this discrepancy? One variable that appears to figure significantly in movements in wealth inequality is the ratio of stock prices to housing prices. Stocks are an asset held primarily by the upper classes, whereas housing is the major asset of the middle classes. If stock prices increase relative to house prices, the share of wealth held by the top wealth groups will rise.
The ratio of stock prices to housing prices (indexed to a value of 15 in 1922 in order to fit the curve onto Figure 6-2) shows a time graph. similar to that of wealth inequality. The ratio more than trebled between 1922 and 1929, corresponding to a tremendous growth in wealth inequality. It fell by half between the 1929 crash and 1933, as wealth inequality also declined, and then it increased by about 20 percent between 1933 and 1939 as the stock market partially recovered, which paralleled a new rise in wealth inequality. The price ratio then fell by almost half between 1939 and 1949 because of a rapid inflation in housing prices, which, in turn, was accompanied by a pronounced: decline in wealth inequality.
The ratio of stock to housing prices more than trebled between 1949 and 1965, and this movement corresponded to a rise in wealth inequality. Between 1965 and 1979, the ratio fell by almost two-thirds, with most of the decline occurring after 1972. Before 1972 the main culprit was rising house prices, but subsequently the principal reason was a stagnating stock market. This period was not surprisingly characterized by a dramatic decline in wealth inequality. Between 1979 and 1989, the price ratio increased by more than half, as the stock market flourished, in line with the sharp increase in wealth inequality that was also recorded over this decade.