International Perspectives on Household Wealth.

Anderson, Kathryn H.

International Perspectives on Household Wealth

Edward N Wolff (ed), Edward Elgar: Cheltenham and Northampton, MA, 2006, 447 pp.

ISBN: 978-1845421168; Price: $180

doi:10.1057/ces.2008.52

Although considerable research has examined the distribution of income over time and in many different countries, we know less about the distribution of wealth. This is because of the paucity of data and the sensitivity of wealth inequality under different institutional arrangements. Essays in this volume describe wealth distributions in seven countries: four on Europe (including East Germany), one on Canada, one on intergenerational transmission in Chile, and five on the United States, all covering years between 1980 and 2000.

The five essays on the United States (by Kennickell; Czajka, Cody & Kasprzyk; Wolff; Ambrosio & Wolff; and Couskova, Juster & Stafford) defined (marketable) wealth as liquid assets minus debts. Liquid assets included the value of owner-occupied housing as well as businesses, financial assets such as stocks, bonds, and mutual funds, transactions accounts, and the cash surrender value of life insurance, 401Ks, and IRAs. Important other sources of wealth, such as the value of social security and private pensions, were omitted. Non-marketable wealth is particularly important to lower income households, who have experienced significant accumulation of social security wealth overtime; this omission biases downward estimates of the value of wealth and may miss an important factor in personal consumption decisions. Four of the essays used data from the Survey of Consumer Finances (SCF) and, with the exception of Czajka, Cody, & Kasprzyk, its high income supplement. The usual summary measures such as the Gini coefficient described the distribution of wealth. Ambrosio and Wolff examined wealth "polarization," which is not inequality but rather the "formation of clusters around local poles," and compared the homogeneity in wealth choices within groups of households to population heterogeneity.

Wolff's paper updated his previous publications. He compared the distribution of income to the distribution of wealth and added an analysis of the 2001 wave of the SCF. Following the 1980-1981 recession, real median household income grew slowly (11%) between 1983 and 1988 but stagnated from 1989-2000. The greatest gains in real household income were at the top of the distribution; income for the top 1% grew 71% from 1982-2000, while for the bottom 80% growth was 25% or less. These differentials led to widening real income inequality; the Gini coefficient increased from 0.48 to 0.56. The racial divide in income also widened over the period. These general patterns were also evident in the wealth distribution. There were fewer wealth-poor households and more wealthy households, especially those with wealth of $1 million or more. The wealthiest 1% increased their net worth by 63% whereas the bottom 40% experienced a decline because of a significant increase in debt.

Kennickell's analysis of the 400 wealthiest showed significant movement in and out of this group, but persistence in their wealth position over time particularly among the wealthiest 100.

Wealth inequality was much higher than income inequality in any year, but the increase in wealth inequality was smaller. The Gini for wealth increased slightly from 0.8 to 0.83. The top 20% of the wealth distribution controlled over 80% of the total net worth. The racial division in wealth widened over the period, and the wealth gap was greater than the income gap by 2001. Stock ownership was increasingly concentrated among wealthy and white households. Financial wealth accumulation followed the same pattern as wealth accumulation, but the patterns was sharper. By 2001, the top 1% of households held 77% of the wealth in stocks whereas the bottom 90% held 74% of the debt and most of the housing. The trend over the period was towards increasing wealth among the elderly in comparison to younger households, who experienced a rapid increase in debt. This life-cycle relationship between age and wealth was also found in Canada and Europe.

Several of the changes observed in the United States were evident in the study of Canada for 1984 and 1999 by Morrissette, Zhang, and Drolet. Employing data from two different household surveys, they measured net worth, including the value of work-related pensions and social-security entitlements (RRSP). Financial wealth increased, but as in the United States, the mean increased more than the median. Only families in the top decile increased their share of total wealth. Inequality increased over time but was significantly lower than in the United States; the Gini coefficient increased from 0.69 in 1984 to 0.73 in 1999, with registered retirement savings plans (RRSPs) and stock ownership accounting for most of the change. Wealth in stocks, business, and other real estate was most unequally distributed. However, among the wealthiest households, RRSPs were the most important wealth component.

Similar trends were documented for Europe, but the extent of inequality over time varied. Hauser and Stein evaluated wealth inequality in Germany over the 1983-1998 period. They included the impact of East-West unification on wealth in the late 1990s. Very high-income households were not included in their study and they omitted the value of business equity, durable goods, and cash from net worth. Over time, the increase in income inequality was largely due to changes in East Germany. German wealth increased over time but was more equally distributed than in North America. Wealth inequality decreased from 1983 to 1993--the Gini fell from 0.68 in 1983 to 0.64 in 1998--but increased significantly after 1993. In contrast to North America, net financial wealth was less unequally distributed than net housing wealth, but inequality ill financial wealth increased slightly over the period while inequality in housing fell.

Average wealth was three times higher in West Germany than in the East, but the distribution of wealth was more unequal in the East than in the West in 1993 and 1998. The intra-German regional inequality gap fell with unification. Inequality in financial wealth had been lower in the East than in the West but rose to about the West level by 1998.

Brandolini, Cannari, D'Alessio, and Faiella examined the distribution of wealth in Italy in the 1990s using household surveys and national accounts. Non-response rates were high, and the wealth of the richest households was underreported, particularly for residences. The adjusted survey data showed growth in mean wealth of 2.7% a year and an increase in wealth inequality; the Gini rose from 0.55 to 0.61. The largest wealth gains were at the top of the distribution with the top S% increasing their share of wealth by over nine percentage points. The distribution of financial wealth became more unequal over time relative to other assets; the Gini for financial wealth increased from 0.66 to 0.81 and was comparable to the measures of financial inequality in North America. Using entropy measures to decompose the increase in inequality, they found that the increase in inequality was primarily the result of widening within-group inequality, not widening group differences. The North and Centre of Italy gained relative to the South and the relative wealth of the college-educated households increased, but widening inequality in wealth within these groups was of greater importance to the change in inequality over time.

The final two papers were essays on Sweden by Klevmarken and Finland by Jantti. The Swedish study presented a serious discussion of public policy changes that affected wealth: deregulation of financial markets and lowered marginal tax rates to encourage savings ill financial instruments. Klevmarken evaluated tax data and a household survey data. The household was defined differently in the two data sources; resident adult children were included in the parents' household in the survey data but comprised separate households in the tax data. The definition of household clearly affected measured inequality. Wealth inequality as reported in the tax data increased; the Gini increased from 0.78 in 1978 to 0.86 in 1997. In the survey data measured inequality was much lower. The change in the distribution of wealth was largely the result of the increased accumulation of financial assets in new private and public pension plans.

In Finland, wealth declined from 1987 to 1994 but then recovered by 1998. The upturn was driven by investment in securities, highly concentrated among the wealthiest households (Gini = 0.98). Accumulation of housing was important to the wealth portfolio of less wealthy households.

The final paper in this volume was, in many respects, the most interesting. Torche and Spilerman's study of Chile used data came--the 2003 Survey of Intergenerational Financial Linkages. Although they could not observe changes in inequality over time, they could compare the wealth position of parents and their adult children, and determine the intergenerational consequences of wealth inequality. Chile had experimented with pension reform and housing policies under both military and socialist regimes. Throughout the 1990s, housing subsidies encouraged households to purchase land and a home, and it encouraged adult children to set up residences separate from their parents. Consequently, the home ownership rate was one of the highest in the world--about 65%, even among the poorest parts of the population. Financial assets were the scarcest, as fewer than 4% of Chilean households own stocks. These assets were highly concentrated among the wealthiest households. The effect of parental wealth on assets of children was positive, working directly through transfers and inheritance and indirectly through income. Because of the housing policy, parental wealth had no direct or indirect impact on home ownership rates. However, parental wealth did affect the value of the home purchased.

From these studies we reach four general conclusions about the global distribution of wealth. First, household wealth increased rapidly from 1980 to 2000 and became more concentrated among the wealthiest households. Inequality in wealth was highest in the United States and Sweden. In contrast, real household income was stagnant, and less concentrate after 1990 in many wealthy countries. Second, evidence from Sweden and Chile illustrate the impact of public policy on our investment choices. Rapid concentration of financial wealth was influenced by financial liberalisation in most countries. Third, wealth accumulation follows a life-cycle pattern, but changes in group membership over time do not explain the important changes in wealth in the 1990s. Finally, wealth transfer is intergenerational; policies that increase wealth inequality today affect wealth inequality tomorrow.

Kathryn H Anderson

Vanderbilt University, Nashville, TN, USA