1995 SCF dataset description

1995 SCF dataset description1995 SCF dataset description

The Survey of Consumer Finances
Insert on page 2 of Federal Reserve Bulletin, January, 1997,
in article, Family finances in the U.S.: Recent Evidence from the Survey of Consumer
Finances, by Arthur B. Kennickell, Martha Starr-McCluer, and
Annika E. Sunde´n. The article is on pages 1-24

The Survey of Consumer Finances (SCF) is a triennial
survey sponsored by the Federal Reserve with the cooperation
of the Department of the Treasury. It is designed to
provide detailed information on U.S. families’ balance
sheets and their use of financial services, as well as on their
pension rights, labor force participation, and demographic
characteristics at the time of the interview. The survey also
collects information on total family income, before taxes,
for the calendar year preceding the survey. The term ‘‘family’’
as it is used here is more comparable to the U.S. Bureau
of the Census definition of ‘‘household’’ than to Census’s
use of ‘‘family,’’ which excludes single people. The appendix
to this article provides a more detailed description of the
SCF, including a more complete definition of this term.
The underlying statistical methodology of the surveys
has been largely unchanged since 1989, and the questionnaires
have been modified only slightly, mostly to reflect
changes in the availability of financial services or in the
financial behavior of families. Thus, the data since that time
are comparable.

The need to measure financial characteristics imposes
special requirements on the survey design. The survey must
provide reliable information both on items that are broadly
distributed in the population—for example, vehicle
ownership—and on items that are highly concentrated in a
relatively small part of the population—for example, investment
real estate. To address this problem, the SCF employs
a dual-frame sample design that includes a standard geographically
based random sample and a special oversample
of relatively wealthy families. Weights are used to combine
information from the two samples for estimates of statistics
for the full population.

This article draws principally upon the final data from the
1992 SCF and preliminary data from the 1995 SCF. Any
differences between figures for 1992 reported here and
those published earlier in the Federal Reserve Bulletin are
attributable to additional statistical processing of the data.
To provide a larger context, some information is also
included from the final version of the 1989 SCF. The SCFs
for 1992 and 1995 were conducted by the National Opinion
Research Center at the University of Chicago (NORC)
between July and December of each survey year. For the
1992 survey, 3,906 families were interviewed, and for the
1995 survey, 4,299 were interviewed.

All dollar figures in this article are adjusted to 1995
dollars using the consumer price index (CPI) for all urban
consumers. Concerns about how accurately the CPI represents
inflation for families have been discussed in the
literature. If, as generally supposed, the index overstates
the true degree of inflation, upward adjustments to past
dollar amounts will appear too large, and trends in positive
dollar amounts will be biased downward. An index sometimes
proposed as an alternative to the CPI is the implicit
price deflator for personal consumption expenditure (PCE),
which is reported as a part of the national income and
product accounts. Over 1989–95, price changes measured
by the CPI and the PCE deflator differed by a relatively
small amount.

To provide a measure of the significance of the developments
discussed in this article, standard errors due to
sampling are given for selected estimates for the 1992
and 1995 data. Space limits have precluded reporting
such figures for all estimates. No standard errors are given
for the 1989 estimates for technical reasons mentioned in
the appendix. Although we do not directly address the
statistical significance of the results, the article highlights
findings that are significant or are interesting in a broader
context.


The following is from pages 22-24

APPENDIX:SURVEY PROCEDURES AND
STATISTICAL MEASURES

Since 1989, the questionnaires for the SCF have
changed only slightly. Generally, changes have been
introduced to gather additional information needed to
understand other data in the survey—for example,
the 1995 SCF introduced a question on uses of funds
for mortgages that were taken out after the time a
primary residence was purchased. Also, the major
aspects of the sample design have been fixed over
this time. Thus, the information obtained by the survey
is comparable over 1989–95.

The survey is intended to provide an adequate
descriptive basis for the analysis of family assets and
liabilities. To address this requirement, the SCF combines
two types of samples. First, a standard multi-stage
area-probability design is selected to provide
good coverage of characteristics, such as home ownership,
that are broadly distributed in the population.
Second, a special list sample is included to over-sample
wealthy families, who hold a disproportionately
large share of such assets as non-corporate
businesses and tax-exempt bonds. This
list sample is drawn from a sample of tax records
made available for this purpose under strict rules
governing confidentiality, the rights of potential
respondents to refuse participation in the survey, and
the types of information from the interview that can
be made generally available. Of the 3,906 completed
interviews in the 1992 SCF, 2,456 families were from
the area-probability sample and 1,450 were from the
list sample; the comparable figures for the 4,299
interviews completed in 1995 are 2,780 families from
the area-probability sample and 1,519 from the list
sample.19

A very important factor in the ability to conduct
surveys is the generosity of the public in giving their
time for an interview. In the 1995 SCF, the average
interview required 90 minutes. However, for some
particularly complicated cases, the amount of time
needed was substantially more than two hours.20

Data for the 1992 and 1995 surveys were collected
by the National Opinion Research Center at the
University of Chicago (NORC) between the months
of June and December in each of the two years. The
great majority of interviews were conducted in person,
although interviewers were allowed to conduct
telephone interviews if that was a better arrangement
for the respondent. In the 1995 survey, one important
change was the introduction of laptop computers for
use in administering the questionnaire. This change
increased the length of the interview somewhat, and
it may also have had some effects on the quality of
information collected.21 Nonetheless, the effects of
the change in the mode of questionnaire administration
appear to be fairly small.

Errors may be introduced into survey results at
many stages. Sampling error, the variability expected
to occur in estimates based on a sample instead of a
census, is a particularly important source of error.
Such error may be reduced either by increasing the
size of the sample or by designing the sample to
reduce important types of variability; the latter course
has been chosen for the SCF. Estimation of sampling
error in the SCF is described further below.
Interviewers may introduce errors, though SCF
interviewers are given lengthy project-specific training
to minimize this problem. In addition, computer
control of the 1995 survey greatly reduced technical
errors made by interviewers. Respondents may introduce
error by understanding a question in a sense
different from that intended by the survey designers.
For the SCF, extensive pretesting and other review of
questions tend to reduce this source of error.
Nonresponse—either complete nonresponse to the
survey or nonresponse to selected items within the
survey—may be another important source of error.
As noted in more detail below, the SCF uses weighting
adjustments to compensate for complete non-response.
To deal with missing information on
individual items, the SCF uses statistical methods to
impute missing data.22

Response rates differ markedly in the two parts
of the SCF sample. In both 1992 and 1995, about
70 percent of families selected for the area-probability
sample actually completed interviews.
The overall response rate in the list sample was about
34 percent. Detailed analysis of the data suggests that
the tendency to refuse participation in an interview is
highly correlated with wealth. The response rates for
both samples are low by the standards of other major
government surveys. However, unlike other surveys,
which almost certainly also have differential non-response
by wealthy families, the SCF sample frame
provides a basis for adjusting for nonresponse by
such families. To provide a measure of the frequency
with which families similar to the sample families
could be expected to be found in the population of
all families, analysis weights are computed for each
case to account for both the systematic properties of
the design and for nonresponse. A major part of
research by SCF staff is devoted to adjustments for
nonresponse through the analysis weights for the
survey.23

For this article, the weights were further adjusted
to decrease the possibility that the results could be
overly affected by a small number of observations.
Such influential observations were detected using a
graphical technique to inspect the underlying data.
Most of the cases found were holders of an unusual
asset or liability or were members of demographic
groups for which such holdings were rare. Trimming
the weights of such cases is likely to make the key
findings in the article more reliable.

To estimate the standard errors due to sampling
that are reported in the main part of this article, a
replication technique was used. Replication methods
draw samples from the set of actual respondents in a
way that incorporates the important dimensions of
the original sample design. In the SCF, a bootstrap
procedure was used to select 1,000 sample replicates,
and a new weight was computed for all the cases in
each of the selected replicates.24

Comparable standard errors are not available for
the 1989 survey. Weights for the 1989 SCF were
computed using an earlier version of the methodology
applied in 1992 and 1995. Although estimates
using these 1989 weights are comparable to estimates
from the later surveys, the difference in the weight
construction is a source of variability in comparisons.
An ongoing project computes weights for the 1989
SCF using exactly the same methodology as that
applied in the later surveys. Corresponding bootstrap
replicates and their weights will also be computed. A
set of tables for the 1989 data reporting the major
detailed estimates presented in this article for 1992
and 1995 will be published in a later issue of the
Federal Reserve Bulletin.

Generally, the survey data correspond well to
external estimates, when such information is available.
Comparisons of SCF data with aggregate data
from the Federal Reserve flow of funds accounts
suggest that when proper adjustments are made to
achieve conceptual comparability, these aggregate
estimates and the SCF estimates for 1989 and 1992
are very close.25 In general, only medians from the
SCF can be compared with those of other surveys
because of the special design of the SCF sample.

The definition of ‘‘family’’ used throughout this
article differs from that typically used in other government
studies. In the SCF, a household unit is
divided into a ‘‘primary economic unit’’ (PEU)—the
family—and everyone else in the household. The
PEU is intended to be the economically dominant
single individual or pair of individuals (who may be
married or living as partners) and all other persons
who are financially dependent on that person or those
persons. In other government studies, for example,
those of the Bureau of the Census, a single individual
is not considered a family. As noted in the main text,
the Census definition of household is closer to the
SCF definition of family. The term ‘‘head’’ used in
this article is an artifact of the organization of the
data and implies no judgment about the structure of
families. In a PEU containing only a single economically
dominant individual, the head is taken to be that
individual. In other PEUs, the head is taken to be the
male in the core couple of the PEU or the older
person in a same-sex couple.

The data used in this article from the 1989 and
1992 SCFs are derived from the final versions of
those surveys. Results reported in this article may
differ in some details from earlier results based on the
preliminary data from those surveys. The 1995 data
used here represent the best estimates at the current
advanced stage of data processing. Data from the
1995 SCF, in a form designed to protect the privacy
of respondents, should be available in the first half of
1997 from the National Technical Information Service,
5285 Port Royal Road, Springfield, VA 22161,
(703) 487-4763.

footnotes
19. The 1992 SCF represents 95.9 million families; the 1995 SCF
represents 99.0 million families.


Note from Sherman Hanna: Sandra Huston pointed out that in Kennickell & Starr-McCluer’s “Changes in Family
Finances from 1989 to 1992: Evidence from the Survey of Consumer Finances”
(Federal Reserve Board Bulletin, October, 1994) — on page 880 in footnote 22 “….The 1989 SCF represents 93.1 million families and the 1992 survery represents 92.9 families.” Apparently this note in the 1994 article was a typo, so be sure to use the number from the 1997 article, that the 1992 SCF represents 95.9 million families (households).


20. The role of interviewers in this effort is normally not sufficiently
recognized. Without the dedication and perseverance of the
project field staff, the survey would not have been possible.
21. For more information on the effects of computer-assisted interviewing
in the 1995 SCF, see Arthur B. Kennickell, ‘‘Using Range
Techniques with CAPI in the 1995 Survey of Consumer Finances’’
22. See Arthur B. Kennickell, ‘‘Imputation of the 1989 Survey of
Consumer Finances: Stochastic Relaxation and Multiple Imputation’’
(Board of Governors of the Federal Reserve System, 1991).
23. For a description of the weighting methodology, see Arthur B.
Kennickell, Douglas A. McManus, and R. Louise Woodburn,
‘‘Weighting Design for the 1992 Survey of Consumer Finances’’
(Board of Governors of the Federal Reserve System, 1996). The
weighting design for the 1995 survey is identical. A review of the
1995 weights will be available later in the year.
24. See Kennickell, McManus, and Woodburn, ‘‘Weighting
Design,’’ for details.
25. For the details of this comparison, see Rochelle Antoniewicz,
‘‘A Comparison of the Household Sector from the Flow of Funds
Accounts and the Survey of Consumer Finances,’’ Review of Income
and Wealth (forthcoming).

This was extracted by Sherman Hanna


Return to SCF Research Group Home Page