Download Chapter 6: Estimates of US Personal Wealth PDF

TitleChapter 6: Estimates of US Personal Wealth
LanguageEnglish
File Size834.3 KB
Total Pages136
Document Text Contents
Page 1

Chapter 6: Estimates of U.S. Personal Wealth

6

Page 2

617

C
ha

pt
er

6


E
st

im
at

es
o

f U
.S

P
er

so
na

l W
ea

lth
C

om
pe

nd
iu

m
o

f F
ed

er
al

E
st

at
e

Ta
x

an
d

P
er

so
na

l W
ea

lth
S

tu
di

es

0

Personal Wealth, 1992-1995

by Barry W. Johnson

Personal Wealth, 1992-1995

n 1992, there were almost 3.7 million adults with
gross assets of at least $600,000 in the United
States. These “top wealthholders” had combined

net worth of almost $5.0 trillion. These individuals
represented just over 2.0 percent of the total U.S. popula-
tion in 1992, yet their wealth accounted for nearly 28.0
percent of total U.S. personal wealth [1]. By 1995, the
number of top wealthholders had increased to 4.1 million
with net worth of almost $5.7 trillion. Some of the growth
between 1992 and 1995, however, can be attributed to the
modest inflation experienced during this period. After
adjusting for inflation, the number of top wealthholders
has actually declined since 1989, the last year for which
SOI estimates were produced.

Background
The distribution and composition of personal wealth in the
United States are topics of great interest among research-
ers and policy planners. Unfortunately, they are difficult
issues to study because, while there are many sources of
data available to examine income trends, data sources on
wealth are scant. The few surveys that attempt to measure
wealth tend to do a poor job of representing the wealthi-
est, and therefore most influential, individuals because of
the relatively small size of this important group. One
exception is the Survey of Consumer Finances (SCF)
sponsored by the Board of Governors of the Federal
Reserve System, a nationwide, household survey that
collects extensive data on assets, debts, income, and
attitudes about finances [2].

Administrative records, specifically the Federal estate
tax return (Form 706), provide an alternative source from
which to study wealth. Detailed descriptions of assets,
debts, and expenses are reported for decedents with total
assets at or above the filing threshold in effect at the time
of death. The estate multiplier technique can be used to
estimate the wealth of living individuals by using data
from these tax returns. The fundamental assumption
underlying this methodology is that estate tax returns,
taken as a whole, represent a random sample, designated
by death, of the living population. Estimates of the wealth
holdings of the living population are derived by applying
a multiplier, based on appropriate mortality rates, to this
sample.

The estate multiplier technique was first used at the
beginning of this century to estimate the wealth of Great

Britain from estate duty records and has been used in
Australia, Italy, the Netherlands, and New Zealand as well
[3]. Horst Mendershausen was the first to apply this
technique in America, producing estimates of U.S. per-
sonal wealth for 1922-46, followed a few years later by
James Smith and Robert Lampman [4]. The Statistics of
Income (SOI) Division has been using the estate multi-
plier technique to estimate the wealth of living individuals
since the 1960’s.

The personal wealth estimates presented in this article
are based on data from Federal estate tax returns. A
decedent’s estate has up to 9 months to file an estate tax
return, and use of a 6-month extension is not uncommon.
It is, therefore, necessary to sample returns filed over a
number of calendar years in order to capture data repre-
sentative of all estate tax decedents dying in a single year.
In the recent past, SOI has combined returns filed over a
3-year period to produce estimates of wealth for a single
year. The estimates presented here for 1992 continue this
practice. The preliminary estimates for 1995, however,
are based on 2 years of filings, adjusted for the remaining,
unfiled returns. This was done in an attempt to provide
more timely estimates; updated 1995 estimates will be
published in the future. One of the strengths of the estate
multiplier technique is the large sample upon which the
estimates are based. The 1992 sample includes nearly
16,000 returns; the 1995 sample contains over 15,000
returns, both considerably larger than samples selected for
other studies at comparable levels of wealth.

Limitations
While the sample size and richness of available data make
this estimation technique attractive, there are limitations to
be kept in mind. The most important is that “estate tax
returns provide a presumably random sample, stratified by
age, not of the total population, but of living persons with
gross estate at or above the filing threshold” [5]. Research
has proven that “individuals who are economically or
socially better off also live longer and are healthier” [6].
Factors such as access to better health services, better diet
and nutrition, fewer risks on the job, and access to better
housing all seem to contribute to this phenomenon [7].
Therefore, determining a mortality rate appropriate to this
sample poses a major challenge. Further, it has been
shown that, while patterns of wealth holding appear quite
robust over a variety of reasonable alternate assumptions
about the multipliers, overall aggregate estimates are
relatively sensitive to the selection of the mortality rates.
This suggests that care should be taken not to give wealth
concentration estimates undue emphasis [8]. (See the

I

Barry W. Johnson is an economist with the Special Studies
Special Projects Section. This article was prepared under the
direction of Michael Alexander, Chief.

Page 68

683

C
h
a
p
te

r
6


E
st

im
a
te

s
o
f
U

.S
P

e
rs

o
n
a
l W

e
a
lth

C
o

m
p
e
n
d
iu

m
o

f
F

e
d
e
ra

l E
st

a
te

T
a
x

a
n
d
P

e
rs

o
n
a
l W

e
a
lth

S
tu

d
ie

s

Personal Wealth, 1998

For females over 85 years of age, the mortality rates
were the same for both groups.

Multipliers
The multipliers (or sample weights) were calculated
as follows:

MULT= 1 / (p ’•r ’•d) where:
p = probability of selection to the estate tax sample,
r = mortality rate,
d= rate differential.
The multipliers ranged from 2 to 18,000, with an

average of about 250. Some additional smoothing of
the multipliers was employed to constrain both tails of
the net worth distribution [20].

Notes and References

[1] See Menchik, Paul (1991), “Economic Status as
a Determinant of Mortality Among Nonwhite
and White Older Males: or, Does Poverty Kill?,”
Institute for Research on Poverty, Discussion
Paper Number 93891.

[2] Scheuren, Fritz (1994), “Historical Perspectives
on IRS Wealth Estimates With a View to Im-
provements,” Compendium of Federal Estate
Tax Data and Personal Wealth Studies,
Department of the Treasury, IRS Publication
1773, p. 358.

[3] Eller, Martha Britton (2001), “Audit Revaluation
of Federal Estate Tax Returns,” Internal
Revenue Service Statistics of Income Bulletin,
Winter 2000-2001, Washington, DC.

[4] Wherever possible, the value of minority discounts
was added back to the individual asset values
used in these estimates. However, inconsisten-
cies in the way that these discounts are reported
on Form 706 limit SOI’s ability to collect these
data. Therefore, it is likely that the estimates
presented here are still somewhat undervalued.

[5] Estimates of the equity value of life insurance
included in total assets were approximated,
based on the face value reported on Federal
estate tax returns and on the decedent’s age. A
ratio of the equity value to the face value was
developed, using data from wealthy respondents
to the 1989, 1992, and 1995 Board of Governors
of the Federal Reserve System's Surveys of

Consumer Finances (SCF). A simple regres-
sion was used to predict the values used in the
Statistics of Income estimates. The same set
of ratios was used for both males and females,
due to a lack of sex-specific data in the SCF.

[6] Estimates of both the total assets and net worth
of the United States are from household
estimates derived from the Board of Governors
of the Federal Reserve System’s Survey of
Consumer Finances (SCF), found in Kennickell,
Arthur, B. (2000), “An Examination of Changes
in the Distribution of Wealth from 1989 to 1998:
Evidence from the Survey of Consumer Fi-
nances,” Board of Governors of the Federal
Reserve System working paper, p. 19.

[7] Marital status estimates for the general popula-
tion, by sex, were obtained from the U.S.
Census Bureau, Current Population Reports,
pp. 20-514.

[8] Estimates of personal wealth for 1995 can be
found in Johnson, Barry W., “Personal Wealth
1995,” Internal Revenue Service Statistics of
Income Bulletin, Winter 1999-2000,
Washingon, DC pp. 59-84.

[9] Closely held stock is stock in a corporation that
is not publicly traded, usually that of a small,
family-owned corporation.

[10] See Modigliani, Franko (1986), “Life Cycle,
Individual Thrift, and the Wealth of Nations,”
American Economic Review, Vol. 68, pp.
547-560.

[11] While the size of the underlying sample of
estate tax returns makes estimates of wealth,
derived using the estate multiplier technique,
fairly robust, estimates of wealth by State can
be subject to significant year-to-year fluctua-
tions for States with relatively small decedent
populations. This is especially true for individu-
als at the extreme tail of the net worth distribu-
tion. For this reason, Table 6 is limited to
individuals with net worth of $20 million or less.

[12] U.S. population data obtained from the U.S.
Bureau of the Census (1999), “Population
Estimates for the U.S., Regions, Divisions, and

Page 69

684

C
h
a
p
te

r
6


E
st

im
a
te

s
o
f
U

.S
P

e
rs

o
n
a
l W

e
a
lth

C
o

m
p
e
n
d
iu

m
o

f
F

e
d
e
ra

l E
st

a
te

T
a
x

a
n
d
P

e
rs

o
n
a
l W

e
a
lth

S
tu

d
ie

s

2

Personal Wealth, 1998

States by 5-year Age Groups and Sex: Annual
Time Series Estimates, July 1, 1990 to July 1,
1998.”

[13] Consumer Price Index (2002), U.S. City
Average for All Items, U.S. Department of
Labor, Bureau of Labor Statistics, Washington,
D.C.

[14] Estimates of the U.S. adult population obtained
from U.S. Bureau of the Census, press release
CB96-88, also published in U.S. Bureau of the
Census, Statistical Abstract of the United
States: 2000 (120th) edition, Washington, DC,
Table 13.

[15] For estimates of U.S. net worth for 1989, 1992,
1995 and 1998, see Kennickell, Arthur, B.,
pp. 16-19.

[16] Although the overall sample of estate tax
returns is large, the number of decedents who
were young (less than 40) or extremely wealthy
(gross assets of $5 million or more) in any given
year varies considerably and is small in com-
parison to their number in the living population.
Because of this, the resulting estimates of
wealth for these two categories of living
individuals would be subject to significant
fluctuations from period to period. To reduce
this variance, the sample is “smoothed” by
including all returns for young or wealthy
decedents filed during the 3-year sample period
without regard to the decedent's year of death.

These segments of the sample are then post-
stratified and reweighted to represent the true
decedent population for the year of interest.
This technique reduces the effect of outliers on
estimates of personal wealth.

[17] Smith, James (1994), “Estimating the Wealth of
Top Wealth-Holders from Estate Tax Returns,”
Compendium of Federal Estate Tax Data
and Personal Wealth Studies, Department of
Treasury, IRS Publication 1773, p. 336.

[18] A more detailed description of this study is
found in A Mortality Study of 1.3. Million
Persons by Demographic, Social, and
Economic Factors: 1979-1985 Follow-up
1992 (1992), U.S. National Longitudinal
Mortality Study, National Institutes of Health,
National Heart, Lung, and Blood Institute, NIH
Publication Number 92-3297.

[19] Mortality data for 1998 were obtained from the
National Center for Health Statistics, Division
of Vital Statistics as reported in the National
Vital Statistics Reports, Volume 48, Number
11, July 24, 2000.

[20] For a more complete methodological discus-
sion, see Johnson, Barry W., “Updating Tech-
niques for Estimating Wealth from Federal
Estate Tax Returns,” 1997 Proceedings of
the American Statistical Association, Section
on Business and Economic Statistics.

SOURCE: IRS, Statistics of Income Bulletin, Winter
2002-2003, Publication 1136 (Rev. 4-2003).

Page 135

750

C
h
a
p
te

r
6


E
st

im
a
te

s
o
f
U

.S
P

e
rs

o
n
a
l W

e
a
lth

C
o

m
p
e
n
d
iu

m
o

f
F

e
d
e
ra

l E
st

a
te

T
a
x

a
n
d
P

e
rs

o
n
a
l W

e
a
lth

S
tu

d
ie

s

Personal Wealth, 2004
Statistics of Income Bulletin | Fall 2008

[All figures are estimates based on samples—numbers are in thousands, money amounts are in millions of dollars]

Number Amount Number Amount Number Amount Number Amount

(25) (26) (27) (28) (29) (30) (31) (32)

Total 376 43,613 1,102 223,274 853 200,790 247 93,509

Under 50 87 8,440 269 45,727 211 59,997 62 21,884

50 under 65 133 16,158 387 63,868 297 61,975 85 34,143

65 under 75 74 6,914 202 35,707 149 36,062 46 17,318

75 under 85 55 8,977 165 53,459 134 28,644 40 14,374
85 and older 27 3,124 79 24,513 61 14,113 14 5,790

Number Amount Number Amount Number Amount Number Amount

(33) (34) (35) (36) (37) (38) (39) (40)

Total 314 41,766 213 183,863 103 121,984 179 151,363

Under 50 75 10,654 62 59,753 11 14,440 43 30,229

50 under 65 127 16,926 78 62,657 38 53,246 61 81,151

65 under 75 53 6,907 45 24,239 30 30,252 42 18,829

75 under 85 43 5,348 21 30,188 16 17,526 23 13,722
85 and older 15 1,931 7 7,026 7 6,519 10 7,431

Number Amount Number Amount Number Amount

(41) (42) (43) (44) (45) (46)

Total 840 331,679 96 24,920 1,059 111,530

Under 50 220 79,313 17 1,847 271 39,267

50 under 65 330 144,566 36 8,515 376 39,213

65 under 75 156 62,698 25 2,662 190 14,655

75 under 85 106 37,538 12 5,081 155 12,283

85 and older 28 7,564 6 6,815 68 6,111

[1] Mutual funds with a single investment objective are grouped with similar direct investments in this table.

[2] Includes individual retirement accounts, annuities, and self-employed retirement or Keogh plans.

NOTE: Detail may not add to total due to rounding.

Table 5. Personal Wealth 2004: Female Top Wealth Holders with Gross Assets of $1.5 Million or More,
Type of Property by Age—Continued

Diversified mutual funds [1]

Limited partnerships

Retirement assets [2] Art

Cash Cash management accounts Mortgages and notes
Age

Age

Other assets

Cash value life insurance Noncorporate business assets

Age

Farm assets

Page 136

751

C
h
a
p
te

r
6


E
st

im
a
te

s
o
f
U

.S
P

e
rs

o
n
a
l W

e
a
lth

C
o

m
p
e
n
d
iu

m
o

f
F

e
d
e
ra

l E
st

a
te

T
a
x

a
n
d
P

e
rs

o
n
a
l W

e
a
lth

S
tu

d
ie

s

Personal Wealth, 2004
Statistics of Income Bulletin | Fall 2008

[All figures are estimates based on samples—numbers are in thousands, money amounts are in millions of dollars]

Number Amount Number Amount Number Amount Number Amount
(1) (2) (3) (4) (5) (6) (7) (8)

Total 2,196 9,721,133 2,194 6,622,433 1,974 2,277,191 2,042 1,441,099
Alabama 18 79,123 18 53,629 17 17,512 18 12,646
Alaska 1 4,776 1 3,384 1 897 1 682
Arizona 36 139,861 36 96,570 33 34,175 34 16,737
Arkansas 11 94,704 11 79,760 10 8,075 11 14,051
California 428 1,793,642 427 996,853 404 691,416 405 247,481
Colorado 32 163,324 32 98,430 29 51,571 31 27,349
Connecticut 47 197,801 47 129,299 41 63,737 45 20,351
Delaware 8 30,923 8 19,418 6 8,121 7 5,007
District of Columbia 7 27,850 7 16,495 7 8,416 7 4,231
Florida 199 904,014 198 612,124 177 221,033 188 138,936
Georgia 56 270,677 56 211,199 51 60,494 54 41,528
Hawaii 7 22,552 6 11,903 7 10,647 5 1,825
Idaho 5 23,982 5 17,509 5 4,055 5 4,187
Illinois 101 476,354 101 349,822 83 77,028 90 68,659
Indiana 32 112,272 32 87,147 26 13,344 29 14,939
Iowa 18 55,332 18 34,395 14 5,635 18 18,996
Kansas 21 65,084 21 45,121 18 8,376 21 13,388
Kentucky 18 65,404 18 48,258 17 8,018 18 11,347
Louisiana 22 92,315 22 72,653 22 15,543 22 10,383
Maine 8 35,173 8 28,677 7 5,950 7 1,928
Maryland 50 191,279 50 134,922 47 37,892 45 27,589
Massachusetts 83 335,482 83 237,389 75 82,248 75 34,079
Michigan 47 261,085 47 192,736 39 34,803 40 50,592
Minnesota 33 135,682 33 92,618 31 25,875 32 23,032
Mississippi 8 61,786 8 33,608 8 5,238 8 23,790
Missouri 33 115,716 33 91,254 26 17,673 32 13,429
Montana 7 23,966 7 16,515 6 4,420 6 3,490
Nebraska 13 83,265 13 68,620 11 6,396 13 11,969
Nevada 15 80,768 15 54,894 14 22,132 15 12,959
New Hampshire 7 27,342 7 18,563 6 7,245 5 2,677
New Jersey 79 324,712 79 219,677 72 79,200 67 40,264
New Mexico 9 28,107 9 18,230 8 4,957 9 6,422
New York 168 942,812 168 636,244 147 218,876 141 143,601
North Carolina 59 223,408 59 171,845 53 41,138 54 25,721
North Dakota 1 3,988 1 2,944 1 777 1 413
Ohio 61 228,532 61 182,596 50 30,049 55 23,694
Oklahoma 17 58,554 17 45,444 14 6,217 17 9,117
Oregon 15 61,328 15 42,631 14 13,685 14 8,515
Pennsylvania 86 399,312 85 293,609 73 70,258 74 47,388
Rhode Island 8 30,782 8 20,882 8 8,124 7 3,533
South Carolina 14 67,856 14 43,678 13 17,713 14 10,988
South Dakota 6 18,850 6 14,181 5 1,433 5 3,949
Tennessee 25 100,778 25 75,826 24 17,670 24 12,433
Texas 108 492,663 108 330,457 102 59,259 107 134,159
Utah 8 52,674 8 43,204 7 5,128 8 8,218
Vermont 4 20,584 4 12,864 4 3,944 4 4,173
Virginia 59 223,984 59 150,855 55 51,348 55 33,420
Washington 50 180,008 50 120,362 46 50,261 49 21,468
West Virginia 12 28,415 12 21,937 10 5,839 12 3,853
Wisconsin 26 127,515 26 103,720 23 22,016 24 10,230
Wyoming 5 106,698 5 97,214 4 3,109 5 9,987
Other areas [3] 5 28,042 5 20,270 5 8,227 5 1,294

[3] Includes U.S. territories and possessions.

NOTE: Detail may not add to total due to rounding.

State of residence

[1] While the size of the underlying sample of estate tax returns makes estimates of wealth derived using the estate multiplier technique fairly robust, estimates of wealth by State can be
subject to significant year-to-year fluctuations. This is especially true for individuals at the extreme tail of the net worth distribution and for States with relatively small decedent
populations.

[2] Includes all stocks, bonds, mutual funds, cash, and cash management accounts.

Table 6. Top Wealth Holders with Net Worth of $1.5 Million or More, Net Worth and Selected Assets, by
State of Residence, 2004 [1]

All other assetsNet worth Financial assets [2] All real estate

Similer Documents