STUDY

Intergenerational Transfers

The MetLife Study of

Inheritance and Wealth Transfer
to Baby Boomers
A Study by the Center for Retirement
Research at Boston College for the
MetLife Mature Market Institute
December 2010

Mature Market
I N S T I T U T E
The MetLife Mature Market Institute®
The MetLife Mature Market Institute is MetLife’s center of expertise in aging, longevity and the
generations and is a recognized thought leader by business, the media, opinion leaders and the
public. The Institute’s groundbreaking research, insights, strategic partnerships and consumer
education expand the knowledge and choices for those in, approaching or working with the
40+ market.

The Institute supports MetLife’s long-standing commitment to identifying emerging issues

and innovative solutions for the challenges of life. MetLife, Inc. is a leading global provider of
insurance, annuities and employee benefit programs, serving 90 million customers in over 60
countries. Through its subsidiaries and affiliates, MetLife holds leading market positions in the
United States, Japan, Latin America, Asia Pacific, Europe and the Middle East.

For more information, please visit: www.MatureMarketInstitute.com.

57 Greens Farms Road

Westport, CT 06880
(203) 221-6580
[email protected]

All of the authors are with the Center for Retirement Research at Boston College (CRR).
Alicia H. Munnell is director of the CRR and the Peter F. Drucker Professor of Management
Sciences at Boston College’s Carroll School of Management. Anthony Webb is a research
economist. Zhenya Karamcheva is a research associate. Andrew Eschtruth is associate director
for external relations.

• Not A Deposit • Not FDIC-Insured • Not Insured By Any Federal Government Agency
• Not Guaranteed By Any Bank Or Credit Union • May Go Down In Value

Text © 2010 Trustees of Boston College, Center for Retirement Research. All rights reserved.
 Table of Contents

2 Executive Summary
2 Key Findings

4 Introduction

5 How Much Will Boomers Receive in Aggregate?

7 How Do Boomers’ Receipts Compare with Those of Previous Birth Cohorts?

8 How Much Does the Inclusion of Inter-vivos Gifts Affect the Numbers?

9 Who Receives How Much, When, and from Whom?

14 How Much Might Boomers Leave to Succeeding Generations?

15 How Has the Recession Affected Prospective Inheritances?

15 Approach 1: How Bequests Changed in the Previous Downturn
17 Approach 2: How Much Bequests Could Change Due to the Current Downturn

19 Implications

20 References

22 Appendix

24 Endnotes
Executive Summary
One potential monetary source that could improve the retirement outlook for
Baby Boomers is inheritances. This study, authored by the Center for Retirement
Research at Boston College for the MetLife Mature Market Institute, quantifies how
much Baby Boomers can expect to inherit over their lifetimes and the distribution
of receipts by household type.

Our best estimate is that Boomers will inherit $8.4 trillion. Of this amount, $2.4
trillion has already been received, while the remaining $6.0 trillion is anticipated
and, therefore, subject to significant uncertainty. The estimates are based on data
obtained before the economic crisis, so our analysis explores how the collapse
in the stock and housing markets might affect the picture. Evidence from the
previous economic crisis in the early 2000s suggests only a temporary reduction in
prospective inheritances, which will be reversed as the economy recovers. However,
given the severity of the recent crisis, we also looked at how inheritances would
be affected if they were to drop proportionately with the decline in housing and
stock values between 2007 and 2010. In this case, anticipated inheritances would
fall 13% — from $6.0 trillion to $5.2 trillion.

We estimate that two-thirds of Boomer households will receive some inheritance

over their lifetimes, with a median amount of $64,000. Although the incidence
of receipt increases with income, 50% or more of households in all income
deciles will eventually receive an inheritance. Among those receiving inheri-
tances, high-wealth households can expect to receive substantially more — an
average of $1.5 million for those in the top wealth decile compared with
$27,000 for those in the bottom decile. However, inheritance receipts have a
proportionately larger impact on the current wealth of low-wealth households.

Regardless of the anticipated amount, any prospective inheritance is uncertain.

Parents or grandparents who expect to leave a bequest may revise their plans
based on fluctuations in their asset values. Or they may exhaust their wealth as
a result of medical costs or long lifespans. In short, Boomer households should
not count on an anticipated inheritance to eliminate the need for increased
retirement saving.

Key Findings
• Over their lifetimes, Baby Boomers are estimated to receive inheritances of
2 $8.4 trillion — with only $2.4 trillion already received and the remaining
$6 trillion anticipated in the future.

• The $8.4 trillion figure is based on 2007 data, which predate the economic
crisis. The impact of the crisis is highly uncertain. Evidence from the previous
market crash suggests little change as the economy recovers, but an alternative
 assumption is that prospective inheritances will fall proportionately with
declining asset values, which would cause a drop of $800 billion or 13%.

• While most transfers occur when parents or grandparents die and leave money
to the younger generation, some transfers occur while the older generations are
still alive. Including these inter-vivos gifts would increase our estimate of total
past and prospective transfers to the Baby Boomers from $8.4 to $11.6 trillion.

• Most Boomers will receive their inheritances in late middle age, reflecting a
pattern in which wealth passes from parents to children on the death of the
surviving parent. To date, the overwhelming majority of the Boomers’ inheri-
tances have been received from parents (63% of inheritances and 74% of
dollars), with grandparents as the second most common source. Few Boomers
now have living grandparents, but the majority has at least one living parent.

• Although only 17% of Boomers had received an inheritance by 2007, two-

thirds will eventually receive one. The incidence of receipt increases with
income, but 50% or more of households in all income deciles will eventually
receive an inheritance.

• Among those receiving an inheritance over their lifetimes, the median amount
is $64,000. The distribution of inheritances is highly unequal. Conditional on
receiving anything, the mean amount received over a lifetime is $1.5 million
for households in the top wealth decile, compared with $27,000 for those in
the bottom. But even within wealth deciles, the distribution of receipts is
highly unequal, and the medians for the top and bottom deciles are $335,000
and $8,000, respectively.

• Though high-wealth households receive much larger inheritances in dollar

terms, these amounts represent a smaller share of their wealth — 22% for
those in the top decile compared to 64% for those in the second-to-bottom
decile.

• Considering only past inheritances, the median amount received by Boomers

by 2007 — adjusted for inflation — is about the same as that received by the
preceding 1927–1945 birth cohort at the same ages.

• Although inheritances are important in aggregate, the timing and amount of

the receipts of any particular household are highly uncertain. Wealth may be
consumed by medical and long-term care costs, or simply as a result of living 3

longer than expected. Wealth is also subject to fluctuations in the stock and
housing markets. So an anticipated inheritance is not a substitute for adequate
saving for retirement.

> Intergenerational Transfers

Introduction
Due to a changing retirement landscape, many Baby Boomers are likely to
have insufficient resources for a secure retirement.1 One potential source of
income that could improve their situation is intergenerational transfers, such
as inheritances. Although the extent to which such transfers contribute to
household wealth has been the subject of vigorous debate, it is clear that they
are important in aggregate, and economically significant for some households.2

This study, authored by the Center for Retirement Research at Boston College
for the MetLife Mature Market Institute, quantifies the aggregate amount that
Baby Boomers, those individuals born between 1946 and 1964, can expect to
receive over their lifetimes, and the distribution of past and prospective receipts
by household type.

The discussion is organized as follows. The first section quantifies the aggregate
amount that Boomers will receive over their lifetimes and reconciles our findings
to those of previous research. The second section compares Boomers’ receipts
with those of previous birth cohorts. The third section considers the potential
impact of inter-vivos transfers — transfers made during the donor’s lifetime. The
fourth section investigates who will receive how much, when, and from whom.
The fifth section considers how much Boomers might in turn transfer to their
children and parents. The sixth section considers the impact of the crises in the
stock and housing markets on intergenerational transfers. The final section
concludes that, while intergenerational transfers will augment the resources of
aging Baby Boomers, they will be insufficient to ensure secure retirements. The
study contains an appendix describing the data and explaining the methodology.

4
 How Much Will Boomers Receive in Aggregate?
To answer this question, we use data from the Survey of Consumer Finances (SCF)
and the Health and Retirement Study (HRS). The SCF is a triennial survey that
oversamples wealthy households; the latest available data are for 2007. When
re-weighted, it is representative of the U.S. population. SCF participants are asked
about past receipts of inheritances and of inter-vivos gifts. They are also asked
whether they expect a substantial inheritance or transfer of assets in the future,
and the anticipated amount. But “substantial” and “expect” are left undefined,
and only 16% of households answer in the affirmative. For data on prospective
inheritance receipts, we therefore turn to the HRS, a nationally representative
panel of individuals born before 1954 and their spouses of any age. In 2006,
individuals were asked to estimate the probability of receiving an inheritance in
the next 10 years, and the likely amount. We convert these 10-year forecasts into
lifetime probabilities, and impute probabilities and amounts to SCF households
born between 1946 and 1964, to obtain a complete picture of past and prospec-
tive inheritance receipts for the Baby Boomers.3 Unfortunately, they were not
asked about prospective inter-vivos transfers.

Inheritances can be received at various ages, and a dollar received at a younger

age is worth more in present value terms than a dollar received later in life.
Therefore, the key question is what interest rate to use for compounding or
discounting the amounts. One approach, taken by Havens and Schervish (1999),
is not to apply any time discounting at all. This method places an equal value
on an inflation-adjusted dollar, regardless of the date of receipt. At the other
extreme, one could discount receipts at the rate of return on financial assets.
This approach produces very large present values for inheritances received in the
distant past and very small present values for those the household anticipates
receiving in the distant future. We follow Brown and Weisbenner (2004) and use
a real discount rate of 3% in our base case, approximating to both the long-term
rate of return on high-grade corporate bonds and commonly used estimates of
the rate of time preference. But to facilitate comparison with Havens and
Schervish (1999), we also report calculations using a zero percent discount rate.
As our objective is to value past and prospective inheritance receipts as of a
current date, we discount and compound all amounts to 2009.4

> Intergenerational Transfers

How Much Will Boomers Receive in Aggregate?

Table 1 reports estimates of aggregate and per household receipts. At a 3%

interest rate, total, past, and prospective receipts amount to $8.4, $2.4, and
$6.0 trillion, respectively. At a zero-percent interest rate, less weight is attached
to past and more weight to prospective receipts. Total, past, and prospective
receipts then amount to $9.2, $1.5, and $7.7 trillion, respectively. The comparable
data per household show similar patterns. In addition, the mean amounts —
except for past receipts — are more than four times the size of the medians,
as those near the top of the distribution receive much larger inheritances in
dollar terms.

Table 1: Aggregate, Mean, and Median Inheritance Receipts, in 2009 Dollars

0% Discount Rate 3% Discount Rate

Amount per Amount per
Total Total
Household Household
Amount Amount
(for recipients) (for recipients)
(trillions) (trillions)
Mean Median Mean Median
Past 1.5 53,315 54,925 2.4 84,494 78,027
Prospective 7.7 266,178 61,235 6.0 207,861 46,926
Total 9.2 319,493 57,932 8.4 292,355 63,875

Note: The medians report the middle inheritance for households with a non-zero inheritance receipt
in the corresponding category.

Source: Authors’ calculations based on U.S. Board of Governors of the Federal Reserve System, Survey
of Consumer Finances, 2007; and University of Michigan, Health and Retirement Study, 2006.

The choice of discount rate has little effect on the 2009 present value of the
sum of past and prospective inheritance receipts. At a higher discount rate,
the increase in the present value of past inheritance receipts is offset by the
reduction in the present value of prospective receipts. To put the $8.4 trillion
in context, our analysis of SCF data shows that, in 2007, total household wealth
for all ages amounted to $65.9 trillion in 2009 dollars.5

The aggregate numbers above are consistent with the findings of previous
research, after adjusting for inflation. Using a zero-percent discount rate,
Havens and Schervish (1999) estimated that during the period 1998–2052,
bequests to heirs of all ages might amount to upward of $24 trillion 1998 dollars
6
and could conceivably be as high as $65 trillion. Havens and Schervish (2003)
estimated that Boomers may well receive $7.2 trillion of the $24 trillion. The $7.2
trillion equates to $9.5 trillion in 2009 dollars. Eliminating inheritances
received prior to 1998 reduces our zero-percent discount rate estimate of $9.2
trillion to $8.4 trillion. Our estimate is thus of the same order of magnitude as
that of Havens and Schervish (2003).6
 How Do Boomers’ Receipts Compare with
Those of Previous Birth Cohorts?
To answer this question, we compare the amounts the Boomers had received
by 2007, when they were 43 to 61 years old, with the amounts that households
born between 1927 and 1945 had received by 1989, when they were also 43
to 61 years old. The 1989 amounts are restated in 2009 dollars, and amounts
received in years prior to 1989 are compounded using a 3% real interest rate.
No comparable 1989 data are available for prospective inheritance receipts.

By 1989, the 1927–1945 birth cohort had received $1.5 trillion, compared with the
$2.4 trillion received by the 1946–1964 birth cohort by 2007 (see Table 2). But the
Boomer birth cohort was nearly 70% larger than the 1927–1945 birth cohort, and
the median present value amount received per household was only slightly larger
in 2009 dollars.7 This result is at first glance surprising, given the dramatic increase
in household wealth during the post-war period. Part of the explanation is reduc-
tions in mortality. The majority of transfers from parents to children occur on the
death of the surviving spouse. For households born 1927–1945, 23% had received
an inheritance by 1989; the comparable figure for those born 1946–1964 was 17%
in 2007. This gap reflects different mortality patterns of their parents’ generation;
25% of households born 1927–1945 had no surviving parent in 1989 — and thus
were candidates for having received substantial inheritances — whereas 21% of
those born 1946–1964 had no surviving parent in 2007. Reduced mortality may
also reduce the amount that individuals plan to bequeath as they require a larger
pile of assets to cover their own needs over a lengthier retirement period. An
additional explanation is that bequests passing to the 1946–1964 birth cohort had
to be shared among a larger number of siblings.

Table 2: Inheritance Receipts and Demographics of Households

with Heads Age 43–61

Item 1989 2007

Total inheritance receipts to date (trillions of dollars) 1.5 2.4

Median inheritance, conditional on receiving inheritance 71,088 78,027
Percent of households that have received inheritance to date 23% 17%
Percent of households without living parents 25% 21%
Average number of siblings for household head 2.7 2.9

Note: All amounts are in 2009 dollars.

Source: Authors’ calculations based on the 1989 and 2007 SCF. 7

> Intergenerational Transfers

How Much Does the Inclusion of Inter-vivos
Gifts Affect the Numbers?
Estimates of inter-vivos transfers can be constructed from the responses of either
recipients or transferors. Using data from the 1986 SCF, Gale and Scholz (1994)
compared responses of recipients and transferors. They found that the use of
data collected from recipients yielded estimates of the participation rate and
aggregate amount transferred that were substantially lower than those based
upon data collected from transferors. This pattern suggests a greater willingness
on the part of respondents to report transfers they have made rather than
transfers they have received. The reasons are likely related to psychology; givers
tend to account for their gifts more carefully because it improves their self
image, while recipients tend to underestimate the gifts they have received
because they do not want to think of themselves as financially dependent.
As a result, estimates constructed from transferor data might be more reliable.

However, while transferor data may be reported more accurately, they suffer from
a different limitation — it is hard to determine the recipients of the transfers.
This limitation makes it difficult to construct a nationally representative sample
of recipients. Our analysis required a sample that is representative of the whole
Boomer generation. Therefore, we used recipient data for our estimates.

Neither recipients nor transferors are asked specifically about college tuition
payments. It seems likely that this type of transfer will be substantially under-
reported by both recipients and transferors.

When inter-vivos transfers are included, transfers received to date increase by

38%, from $2.4 trillion to $3.4 trillion. If inter-vivos transfers represent the same
proportion of prospective inheritances, then the lifetime total increases from
$8.4 to $11.6 trillion.

8
 Who Receives How Much, When,
and from Whom?
Past inheritance receipts are highly unequally distributed, with just 17% of
households having received anything. But, in 2007, 58% of Boomer households
had at least one living parent,8 and analyses of amounts received to date there-
fore substantially understate the proportion of households that will eventually
receive an inheritance. Taking into account both past and future inheritances, we
estimate that two-thirds of Boomer households will receive some inheritance.
This figure may sound surprisingly high. To verify our results, we compared data
from both the SCF and HRS on past inheritances for Early Boomers. The results
from the two datasets are comparable with about 20% of households having
already received an inheritance by 2007 (SCF) or 2006 (HRS). To calculate future
inheritances, we relied exclusively on the HRS due to data limitations of the SCF.
In both 2004 and 2006, about 50% of Early Boomers report a positive probability
of receiving an inheritance in the next 10 years. Combining the past and future
figures suggests that our estimate that two-thirds of households will eventually
receive an inheritance is reasonable.

Figure 1 shows the distribution of lifetime inheritance receipts. About one-third

of Boomer households expect to receive no inheritance over their lifetime.
About three-quarters expect either no inheritance or less than $100,000, and
97% expect either no inheritance or less than $1 million.9

Figure 1: Distribution of Sum of Past and Prospective Inheritance Receipts

to Boomers
$10,000,000

$1,000,000
Amount of Inheritance

$100,000

$10,000

$1,000
0% 10% 20% 30% 40% 50% 60% 70% 80% 90%
Percent of Households

Source: Authors’ calculations based on 2007 SCF and 2006 HRS.

> Intergenerational Transfers
 Who Receives How Much, When, and from Whom?

Figure 2 analyzes the percentage of Boomer households that have either

received or anticipate receiving an inheritance, by income decile. Although the
incidence of receipt increases with income, 50% or more of households in all
income deciles will eventually receive an inheritance.

Figure 2: Percent of Boomer Households Who Will Receive an Inheritance Over

Their Lifetime, by Income Decile

90%
80%

70%
60%
50%

40%
30%
20%

10%

0%
1 2 3 4 5 6 7 8 9 10
Income decile

Source: Authors’ calculations based on 2007 SCF and 2006 HRS.

Figure 3 shows mean lifetime receipt, conditional on receiving anything, by

wealth decile.10 High-wealth households receive substantially larger inheritances
— an average of $1.5 million for those in the top wealth decile, compared with
$27,000 for those in the bottom wealth decile. Even within wealth deciles, inher-
itances are highly unequally distributed; the medians for the top and bottom
deciles are $335,000 and $8,000, respectively.

10
Figure 3 also shows the mean amount received as a percent of mean household
wealth for the relevant wealth decile. On average, inheritance receipts represent
a smaller percentage of current wealth for high-wealth households — 22% for
those households in the top wealth decile, compared with 64% for those house-
holds in the second-to-bottom wealth decile.11

Figure 3: Mean Lifetime Receipt and Mean Lifetime Receipt As a Percent

of Wealth for Boomers Receiving an Inheritance, by Wealth Decile

1,600 80%

1,400 70%

1,200
Thousands of (2009) dollars

60%

Percentage of wealth
1,000 50%

800 40%

600 30%

400 20%

200 10%

0 0%
1 2 3 4 5 6 7 8 9 10
Wealth decile

Average lifetime inheritance Inheritance as a percent of wealth

Source: Authors’ calculations based on 2007 SCF and 2006 HRS.

11

> Intergenerational Transfers

 Who Receives How Much, When, and from Whom?

Figure 4 shows the average lifetime receipt, by wealth decile. In contrast to

Figure 3, which shows the average amount, conditional on receiving anything,
this figure shows unconditional means. Households in the top wealth decile
receive dramatically more than those in the bottom decile — an average of
$1.1 million, compared with $9,000. But inheritances represent a smaller
percentage of the current wealth of high-wealth households — 20% for those
in the top wealth decile, compared with 50% for those in the second-to-bottom
wealth decile.

Figure 4: Mean Inheritance Receipt for All Boomers, by Wealth Decile

1,200 60%

1,000 50%
Thousands of (2009) dollars

800 40%

600 30%

400 20%

200 10%

0 0%
1 2 3 4 5 6 7 8 9 10
Wealth decile

Average lifetime inheritance Inheritance as a percent of wealth

Source: Authors’ calculations based on 2007 SCF and 2006 HRS.

Figure 5 shows the percent of Boomers receiving an inheritance by five-year age

group.12 Less than 1% of Boomers received an inheritance between age 20 and
24. But the probability of receipt increases with age, reflecting a pattern in
which most inheritances are received from parents, passing on the death of the
surviving parent, with Boomers now approaching the ages at which many such
deaths are occurring. So 13.7% of those Boomers who had attained age 59 by
12
2007 received an inheritance between ages 55 and 59.
Figure 5: Percent of Boomers Receiving an Inheritance, by Age at Receipt

15%

12%

9%

6%

3%

0%
20–24 25–29 30–34 35–39 40–44 45–49 50–54 55–59

Source: Authors’ calculations based on 2007 SCF.

Table 3 analyzes past inheritance receipts by source. To date, the overwhelming

majority of the Boomers’ inheritances have been received from parents, with
grandparents and aunts and uncles being the second and third most common
sources. Few Boomers now have living grandparents, but the majority has at
least one living parent. So parents are likely to be the predominant source of
inheritances in coming years.

Table 3: Distribution of Inheritances Received by Boomers, by Source

Percent of total Percent of total

Source
number of inheritances amount of inheritances

Parents 63.2% 74.1%

Grandparents 20.4% 16.5%
Aunt/uncle 11.8% 4.7%
Other 4.6% 4.8%
Total 100.0% 100.0%

Source: Authors’ calculations based on the 2007 SCF.

13

> Intergenerational Transfers

 How Much Might Boomers Leave to
Succeeding Generations?
The Boomer Generation is currently 46 to 64 years old. Munnell, Webb, and
Golub-Sass (2009) document the inadequate financial preparedness for retirement
of this birth cohort. This cohort also faces rising health care costs, the largely
uninsured cost of long-term care, the risk of job loss, and declines in prospective
returns on their accumulated wealth. Some have yet to face the potentially
substantial cost of paying for their children’s college education and may also be
called upon to contribute to the cost of their grandchildren’s education. The
above factors could lead one to conclude that the Boomers might be less likely
than previous birth cohorts to leave bequests to succeeding generations.

Offsetting the above factors are the effects of the change in the composition of
pension wealth. Defined contribution pension plans, which are rarely annuitized,
have rapidly displaced defined benefit pension plans in the private sector, while
lump-sum options have become more prevalent in the remaining defined benefit
pension plans. Households managing the decumulation of unannuitized wealth
will almost invariably die leaving part of that wealth unconsumed, even if they
lack an operative bequest motive. Munnell et al. (2003) estimate that the increase
in pension lump-sum payments over the period 1992–2004 resulted in substantial
increases in bequests. They also found evidence that the growth in lump-sum pay-
ments might be raising the interest of households in leaving an intended bequest.

Estimating how these offsetting factors will affect Boomer bequest behavior
is difficult given the 20- to 50-year remaining life expectancy of the Boomer
cohort. In addition, considering that a large number of Boomers have not yet
received inheritances themselves, we conclude that their current bequest inten-
tions may provide little information about their future bequest decisions.

14
 How Has the Recession Affected
Prospective Inheritances?
In this section, we consider how the financial crisis might affect Boomers’
prospective inheritance receipts. The crisis has resulted in steep declines in both
housing and stock prices. These declines might affect prospective inheritances in
a number of ways. At one extreme, Boomers’ parents might strive to maintain
pre-crisis consumption, letting the entire burden of the reduction in asset values
fall on inheritances. At the other extreme, they might choose to reduce current
consumption while attempting to leave the same amount of wealth to their
heirs as previously. Or they might decide to make proportionate cutbacks in both
current consumption and bequests.13

To determine the effects of the crisis, we would, ideally, use survey data collected
after the crisis began. The 2008 HRS does not meet our needs for two reasons.
First, most of the households were interviewed before the collapse of Lehman
Brothers in September 2008, which ushered in the most serious phase of the
financial crisis. Second, the 2008 HRS has more limited data on inheritances
than previous years of the survey.14 The SCF is also not an option as the latest
full version of the survey was conducted in 2007, before the crisis, and a special
version conducted in 2009 is not yet available.

Due to these data limitations, we adopt two complementary approaches to

quantifying the impact of the current recession on prospective inheritances. The
first examines the reactions of households to the previous economic downturn
in the early 2000s. The second directly assesses the current downturn by assuming
that inheritances will fall proportionately with the decline in asset values experi-
enced since 2008.

Approach 1: How Bequests Changed in the

Previous Downturn
HRS data allow us to draw inferences from households’ responses to past financial
crises, in particular the decline in stock prices from 2000 to 2003. Since 1994, HRS
participants have been asked to assess the probability of leaving an inheritance
of greater than $10,000. Those participants answering zero are asked to assess
the probability of leaving any bequest. Those assessing the probability at greater
than zero are then asked to assess the probability that they will leave an amount
exceeding $100,000.15

Using HRS data in pooled cross section, we estimate an ordered probit model in 15

which the dependent variable takes the value zero if an individual is certain he
and his spouse will leave no inheritance; one if he attaches a non-zero probability
to leaving an amount greater than zero but is certain he will not leave at least

> Intergenerational Transfers

 How Has the Recession Affected Prospective Inheritances?

$10,000; two if he attaches a non-zero probability to leaving an amount greater

than $10,000 but is certain he will not leave at least $100,000; three if he attaches
a non-zero probability to leaving an amount greater than $100,000; and four if
he is certain he will leave an amount exceeding $100,000.

We control for a variety of socioeconomic characteristics, include a full set of year

dummies, and restrict our analysis to those aged 65 or over at the date of the
interview. We hypothesize that households will be less likely to report that they
anticipate leaving a large inheritance in 2002, when there had been a substantial
decline in the stock market. We further hypothesize that the greatest effect
will be among households with substantial exposure to the stock market. We
therefore interact the year dummies with a variable taking the value one if the
household had $25,000 or more invested in the stock market, zero otherwise.

Table 4 reports the impact of the year dummies and year dummies interacted
with stock ownership on the predicted percent of households falling into five
response categories, holding personal characteristics constant at their means.
So in 2000, the entries in the first row show that a predicted 2.1% of households
with more than $25,000 invested in the stock market would report that they had
a zero percent chance of leaving any inheritance. Just over 20% definitely plan
to leave an inheritance of $100,000 or more.

Table 4: Estimated Distribution of Households by Anticipated Bequest,

Respondents with More than $25,000 in Equity Assets

Certain of Some probability of leaving Certain of

Year leaving $10,000– leaving
$1–$9,999 $100,000+
nothing $99,999 $100,000+

2000 2.1% 2.3% 19.0% 54.1% 22.5%

2002 2.4% 2.6% 20.4% 54.0% 20.5%
2004 2.8% 3.0% 21.8% 53.6% 18.7%
2006 2.4% 2.6% 20.2% 54.0% 20.9%

Source: Authors’ calculations based on the 2000, 2002, 2004, and 2006 HRS.

Although the changes in probabilities are sometimes statistically significant, they

are generally small. For example, the decline in stock prices between 2000 and
16
2004 was associated with only a 3.8 percentage-point decline in the number of
households who were certain of leaving an inheritance of $100,000 or more. Not
only is this decline small but, by 2006, it began to reverse.
Although households with substantial stock portfolios were more likely to antici-
pate leaving a substantial inheritance, there was no statistically significant differ-
ence between their response to the downturn in the stock market between 2000
and 2004 and that of households with little or no exposure to the stock market.

One possible explanation for the small magnitude of the above responses is that
older households plan to reduce consumption in order to preserve the value of
their intended bequest. But we cannot rule out the possibility that households
take time to adjust their expectations in the light of realized investment returns.

Approach 2: How Much Bequests Could

Change Due to the Current Downturn
The previous analysis found only small reductions in anticipated bequests in the
previous downturn. As noted, this finding is consistent with the notion that
older households are willing to sacrifice consumption to maintain their intended
bequests. If so, this result represents one possible extreme. The other extreme is
that older households are not willing to sacrifice their consumption and, instead,
reduce intended bequests by the full amount of the decline in assets that they
experience, an outcome that we deem unlikely. Therefore, our second approach
assumes cutbacks in both consumption and intended bequests.

We assume that anticipated inheritance receipts decrease proportionately with

the declines in stock and house prices from the date of the 2006 HRS interview to
May 2010.16 Respondents are not asked about their parents’ portfolio allocations.
We therefore assume that all persons who might leave inheritances to the
Boomers have the average portfolio allocation for households aged 65 and over,
calculated from SCF data.17 Table 5 shows the composition of total wealth of this
age group as of 2007. Housing comprises 33.1% of the total. We assume that
50% of defined contribution and 80% of mutual fund balances are held in stocks,
so stocks represent 6% of total household wealth. The average date of the HRS
interviews was June 2006. Between June 2006 and June 2010, house prices, as
measured by the S&P/Case-Shiller index, declined by 28.3%.18 The S&P 500 index
declined by 18.9% over the same period.19 Multiplying by the percentages of
household wealth invested in stocks and housing yields a decline in wealth of
13.1%. Applying this percentage reduction to our earlier estimate of prospective
inheritances reduces the estimate from $6.0 to $5.2 trillion.20 Thus, the declines in
17
asset prices will reduce inheritance receipts, but substantial amounts will still pass
to succeeding generations.

> Intergenerational Transfers

 How Has the Recession Affected Prospective Inheritances?

Table 5: Composition of Portfolios of Households Aged 65 and Over, 2007

Asset Class Percent of Total

Net Housing 33.1%

Business Assets 18.3%
Defined Contribution Assets 9.4%
Net Other Nonfinancial Assets 7.9%
Net Financial Assets 31.3%
Transaction Accounts 4.4%
CDs 2.3%
Mutual Funds 6.0%
Stocks 10.1%
Bonds 2.4%
Other 6.1%
Total 100.0%

Source: Authors’ calculations based on 2007 SCF.

18
 Implications
Boomers are likely to receive inheritances totaling upward of $8 trillion over
their lifetimes. Among the two-thirds of Boomer households expected to
receive an inheritance, the median amount is $64,000. Inheritance receipts are
highly unequally distributed, and many households will receive small amounts.
But a substantial minority can expect to receive amounts that will improve
their financial preparedness for retirement, and their ability to pass wealth to
succeeding generations.

These findings have useful implications for Boomers, their financial advisors,
and policymakers concerned about retirement security. First, it is important to
stress that most Boomers have not yet received any inheritance. And the amount
and timing of inheritance receipts is highly uncertain. Even parents who have a
strong desire to leave a bequest may be forced to revise their plans based on
fluctuations in the value of their assets. Or they may exhaust their wealth as a
result of medical and especially long-term care costs. In short, an anticipated
inheritance may not materialize. Even when inheritances do occur, recipients
generally get the money when they are older and the amounts are typically not
large enough to be life-changing. Therefore, Boomer households need to make
many of their key financial decisions before they ever receive any inheritance.
And they should not count on an inheritance to eliminate the need for increased
retirement saving.21

Second, to improve Boomers’ ability to gauge how inheritances might help

them, the subject could be used to generate intra-family discussions about
estate planning. While not everyone will be comfortable engaging in this
subject, those who are able to do so will likely find the opportunity helpful.
A trusted family financial advisor may be useful in this regard as an outside
individual who can help bring both parents and children together to handle
the topic.

Finally, policymakers should recognize that inheritances are not a silver bullet to
achieve retirement security for an aging population, so they should continue to
develop initiatives to boost Americans’ saving and promote longer worklives.
One policy that will not have a major effect on Boomers’ retirement preparedness
is the estate tax. Regardless of whether the estate tax is reinstated at the 2002 or
2009 levels, it will affect only a very small minority of households.

19

> Intergenerational Transfers

 References
Avery, Robert B., and Michael S. Rendall. 1993. “Estimating the Size and
Distribution of Baby Boomers’ Prospective Inheritances.” 1993 Proceedings of the
Social Statistics Section. Alexandria, VA: American Statistical Association: 11-19.

Brown, Jeffrey R., and Scott J. Weisbenner. 2004. “Is a Bird in the Hand Worth
More Than a Bird in the Bush? Intergenerational Transfers and Savings Behavior.”
In Perspectives on the Economics of Aging, edited by David A. Wise, 181-201.
Chicago: University of Chicago Press.

Coe, Norma B., and Anthony Webb. 2009. “Actual and Anticipated Inheritance
Receipts.” Working Paper 2009-32. Chestnut Hill, MA: Center for Retirement
Research at Boston College.

Gale, William G., and John Karl Scholz. 1994. “Intergenerational Transfers and
the Accumulation of Wealth.” Journal of Economic Perspectives 8(4): 145-160.

Havens, John J., and Paul G. Schervish. 1999. “Millionaires and the Millennium:
New Estimates of the Forthcoming Wealth Transfer and the prospects for a
Golden Age of Philanthropy.” Working Paper. Chestnut Hill, MA: Social Welfare
Institute at Boston College.

Havens, John J., and Paul G. Schervish. 2003. “Why the $41 Trillion Wealth
Transfer Estimate is Still Valid: A Review of Challenges and Questions.” Journal
of Gift Planning (7): 11-15, 47-50.

Kessler, Denis, and André Masson. 1989. “Bequest and Wealth Accumulation: Are
Some Pieces of the Puzzle Missing?” Journal of Economic Perspectives (3): 141-52.

Kotlikoff, Laurence J., and Lawrence H. Summers. 1981. “The Role of Intergenera-
tional Transfers in Aggregate Capital Accumulation.” Journal of Political Economy
Vol. 89 pp. 706-732.

Kotlikoff, Laurence J. 1988. “Intergenerational Transfers and Savings.” Journal of

Economic Perspectives 2(2): 41-58.

Modigliani, Franco. 1988. “The Role of Intergenerational Transfers and Life Cycle
Saving in the Accumulation of Wealth.” Journal of Economic Perspectives Spring
Vol. 2 pp. 15-40.

20
Munnell, Alicia H., Annika Sundén, Mauricio Soto, and Catherine Taylor. 2003.
“The Impact of Defined Contribution Plans on Bequests.” Death and Dollars: The
Role of Gifts and Bequests in America, edited by Alicia H. Munnell and Annika
Sundén, 265-316. Washington, DC: Brookings Institution Press.

Munnell, Alicia H., Anthony Webb, and Francesca Golub-Sass. 2009. “The
National Retirement Risk Index: After the Crash.” Issue in Brief 2009-22.
Chestnut Hill, MA: Center for Retirement Research at Boston College.

Vanderhei, Jack and Craig Copeland. 2010. “The EBRI Retirement Readiness
Rating:™ Retirement Income Preparation and Future Prospects.” Issue Brief 344.
Washington, DC: Employee Benefit Research Institute.

21

> Intergenerational Transfers

 Appendix
This appendix describes the questions asked of participants in the Survey
of Consumer Finances (SCF) and the Health and Retirement Study (HRS) and
outlines the methodology used to produce estimates of past and prospective
inheritance receipts.

The financial respondent in each household in the SCF is asked:

Have you or your husband/wife/partner ever received an inheritance or been
given substantial assets in a trust or in some other form?

For the largest three receipts, the respondent is asked the value at the time of
receipt, the date of receipt, whether it was an inheritance, inter-vivos gift, or
something else, and from whom it was received. Respondents are also asked to
state the total of any other receipts, but are not asked to identify dates or sources.

In 2006, but not in 2008, respondents in the HRS were asked:

Not counting anything you might give or leave to each other, what are the
chances that you or your husband/wife/partner will receive an inheritance
during the next 10 years?

Respondents who assess the probability at greater than zero are then asked:
About how large do you expect that inheritance to be?22

Respondents who are unable to provide a precise answer are then invited to
indicate whether the anticipated receipt lies within various ranges.

Assigning prospective inheritance receipts to the SCF households involves seven

steps. We first estimate a probit model on HRS data. The dependent variable
takes the value one if the individual assigns a positive probability to receiving
an inheritance, zero otherwise. Explanatory variables include own and spouse’s
education, ethnicity, log of income, age, number of siblings, indicators for self
and spouse having no living parents, and parents’ ages — conditional on being
alive. We use the parameter estimates to impute a zero or non-zero probability
of receiving an inheritance to Boomer households in the SCF.

We then use hot-deck imputation to assign probabilities of receipt to those SCF

households that have been assigned a non-zero probability of receipt in the
above step. Covariates include marital status, ethnicity, whether both parents
are dead, and number of siblings.

22 We then estimate an OLS regression on those HRS households that assess their
probability of receipt at greater than zero. The dependent variable is the log of
the anticipated dollar amount, and the explanatory variables are those used in
the probit model.
HRS participants are only asked to forecast inheritance receipts up to a 10-year
time horizon. We estimate that in 10 years’ time, 36% of Boomers will still have
at least one surviving parent. So forecasts over a 10-year horizon will substantially
understate lifetime receipts. To obtain estimates of the lifetime probability of
receipt, we multiply the self-reported probabilities assigned to the SCF households
by the reciprocal of the probability that both husband’s and wife’s surviving
parents will die within the next 10 years.23 In some cases, this produces a probability
that exceeds 100%.24 In these cases, we cap the probability at 100% and make a
corresponding adjustment in the anticipated dollar amount.

We then make random draws from the uniform [0,1] distribution, and based on
the values drawn, assign either receipt or non-receipt to each household. Those
assigned receipt of an inheritance are assumed to receive the predicted amount,
plus a draw from a mean zero error term.25

An important assumption underlying the above procedure is that, conditioning

on the socioeconomic characteristics used in the above regressions, prior receipt
of an inheritance does not affect the probability of receiving one in the future.
In theory, prior receipt might be associated with either an increased or decreased
probability of receiving an inheritance in the future. If prior receipt is associated
with an increased probability of receiving an inheritance in the future, our calcu-
lations will understate the extent to which the receipt of substantial inheritances
is concentrated among a small segment of the population. Conversely, if prior
receipt is associated with a reduced probability of subsequent receipt, our calcu-
lations will overstate the concentration of inheritance receipts.

The present value of an inheritance depends on the timing of its receipt.

HRS respondents are not asked when they anticipate receipt. We assume that
all inheritances pass on the death of the surviving parent. The sixth stage
in the calculation involves discounting anticipated receipts by joint parental
life expectancy.

SCF respondents are interviewed in 2007, whereas Health and Retirement Study
participants are interviewed in the latter part of the previous year. In the seventh
and final step, to eliminate double counting of inheritances received between
2006 and 2007, we exclude inheritances received during this period from all
calculations based upon the sum of past and prospective inheritances.26

23

> Intergenerational Transfers

 Endnotes
1 These changes include declining Social Security replacement rates, the shift to 401(k) plans accompanied
by low saving rates, longer lifespans, and rising health care costs. For more details, see Vanderhei and
Copeland (2010); and Munnell, Webb, and Golub-Sass (2009).

2 Kotlikoff and Summers (1981) estimated that 46% to 80% of household wealth is due to inheritances.
In contrast, Modigliani (1988) estimated that only 20% was the result of intergenerational transfers.
Kotlikoff (1988) finds that the difference between the two calculations is largely the result of differences
in the treatment of earnings on inherited wealth.

3 A potential concern is that households may under-report receipts (Kessler and Masson, 1989; and Gale
and Scholz, 1994). However, Brown and Weisbenner (2004) show that aggregate numbers obtained
from self-reported inheritance data are consistent with those obtained from models of transfer flows.
And Coe and Webb (2009) compare households’ estimates of prospective inheritance receipts with
subsequent outcomes and find no evidence of systematic biases.

4 An alternative would be to discount inheritances to a common age, which would decrease the value
of the inheritances of older relative to those of younger households.

5 Household wealth is defined as the sum of financial assets, housing wealth, business assets, and other
nonfinancial wealth exclusive of defined benefit pensions and Social Security assets, net of any debt or
other liabilities.

6 Avery and Rendall (1993) estimate total inheritance flows over the period 1990–2040 at $10 trillion in
1990 dollars. They do not indicate what portion will be received by the Baby Boomers. Their number
can therefore only be compared with the Havens and Schervish (1999) $24 trillion total for all birth
cohorts, which is 140% larger. Our analyses of SCF data show that the total wealth of households aged
65 and over increased by 60% between 1989 and 1998. The Avery and Rendell estimate therefore
appears somewhat low, relative to those of Havens and Schervish (1999) and to our own estimate.

7 In 2007, the Boomer cohort comprised 43.7 million households. In 1989, the 1927–1945 birth cohort
comprised 26 million households.

8 For a married household, we define “having a living parent” as both spouses having at least one
living parent.

9 All the calculations in this section compound past inheritances and discount anticipated inheritances
at a 3% real interest rate.

10 For our definition of wealth, see endnote 5. In addition, when calculating household wealth deciles,
we include the present value of anticipated inheritance receipts.

11 Our emphasis is on the second-to-bottom wealth decile rather than the bottom decile because house-
holds in the bottom decile have so little wealth, even after including inheritances, that inheritances as
a percent of wealth is not a meaningful number.

12 Our calculations are adjusted to reflect the fact that not all Boomers are observed to age 61.

13 This last outcome may not necessarily arise as a result of a desire to leave a bequest. A household that
has a constant relative risk aversion utility function and is lacking a bequest motive would adjust its
consumption proportionately with the reduction in wealth, and would leave a proportionately
smaller bequest.

14 In 2009, a subsample of HRS participants completed an internet questionnaire that included questions
about bequest intentions. These respondents have higher than average socio-economic status.
Therefore, we decided it would be misleading to use these data to draw inferences about the impact
of the financial crisis on bequest intentions.

15 In 2004 and subsequent years, those estimating the probability of leaving $100,000 or more at greater
than zero are asked to assess the probability of leaving $500,000 or more.

24 16 We use the S&P/Case-Shiller house price index and the S&P 500.

17 There has been an approximately proportionate decline in both stock and house prices. So it is unclear
whether the financial crisis has had a greater effect on the wealthy, who hold a larger proportion of
their wealth in stocks, or on the majority of the population, who hold most of their non-pension
wealth in housing.
18 The June 2006 and June 2010 index values were 206.38 and 147.97, respectively.
http://www.standardandpoors.com/indices/main/en/us

19 The S&P 500 closed at 1270.2 on 30 June 2006 and 1030.71 on 30 June 2010.

20 Business assets comprise 18.3% of the wealth of households over 65, and it is likely that the recession
also had a substantial impact on the market value and profit earning potential of these assets.

21 Brown and Weisbenner (2004) find that although past inheritance receipts “crowd out” saving by the
recipient household dollar for dollar, the same is not true of anticipated receipts.

22 We assume that responses are in 2006 dollars, and are not subject to time discounting.

23 As we lack data on the identities of persons from whom the household anticipates receiving an
inheritance, we make the simplifying assumption that all inheritances will be received from the
Boomers’ parents. We further assume that the amount received from a parent or other relative does
not depend on the parent or other relative’s age of death. Households tend to decumulate wealth
as they age, but wealthy individuals have greater average wealth, and give birth at older ages, so
the direction of any bias is unclear.

24 Some respondents will, often with good reason, assess the probability of their parents dying within
10 years at greater or less than that predicted by life tables, in which case our methodology will over-
or understate the lifetime probability of receipt. Others may anticipate receiving inheritances from
more than one source over their lifetime.

25 The error term captures variation in the amounts of anticipated inheritance receipts that is uncorrelated
with the explanatory variables used in our econometric model.

26 Inheritance receipts are coded in five-year intervals. We randomly assign a year of receipt to inheritances
received during the period 2003–2007.

25

> Intergenerational Transfers

 ®

Mature Market
I N S T I T U T E

www.MatureMarketInstitute.com
200 Park Avenue, New York, NY

MMI00165(1210) ©2010 Metropolitan Life Insurance Company L1110144594[exp1111]