Review of Income and Wealth
Series
45,
Number
1,
March
1999
THE IMPORTANCE OF ENTREPRENEURSHIP
FOR
WEALTH
CONCENTRATION AND MOBILITY
BY
VINCENZO QUADRINI
Duke University
The paper conducts an empirical analysis of the importance of entrepreneurship for wealth concen-
tration and mobility using data from the Panel Study of Income Dynamics. The data shows a marked
concentration of wealth in the hands of entrepreneurs which is not merely a consequence of their
higher incomes. The higher saving rates among entrepreneurs is one of the possible explanations for
their higher asset holdings and this hypothesis is supported by the statistical tests conducted in the
paper. The data also shows that entrepreneurs experience greater upward mobility in that they have
a greater probability of moving to higher wealth classes, and this is not only a consequence of their
higher incomes.
As is well known, household wealth is highly concentrated, even more con-
centrated than income. For example Wolff (1995), using data from the Survey of
Consumer Finances, reports that in 1989 the top one percent of households
owned 39 percent of total wealth. Yet the reason why some families-those at
the top of the distribution-accumulate such a high level of wealth is at present
unknown and it constitutes a puzzle that a large class of calibrated models is
not able to capture as discussed in Quadrini and
Rios-Ru11 (1997). The question
addressed in this paper is whether entrepreneurship plays an important role in
generating such a high concentration of wealth.
Using data from the Panel Study of Income Dynamics, the paper documents
the main differences in asset holdings and wealth mobility between entrepreneurs
and workers, where "entrepreneurs" are defined as families owning their own
business and "workers" are defined as all other families. The data analysis shows
that there is a marked concentration of wealth in the hands of entrepreneurs
and that entrepreneurs experience greater upward mobility than workers. These
differences in asset holdings and mobility are not merely accounted for by higher
entrepreneurial incomes, as entrepreneurs have higher wealth-income ratios than
workers, and they experience greater upward mobility in the wealth-income ratio
as well.
The fact that business families own more wealth has been interpreted as
evidence of the existence of borrowing constraints: that is, the ownership of a
business can only in part be financed with external funds, and therefore, only
Note:
I
would like to thank Hilary Appel, Thomas Cooley, Michael Greenacre, Boyan Jovanovic,
Per Krusell, Jost-Victor Rios-Rull and Kenneth Wolpin for helpful suggestions on earlier versions of
this paper.
I
am also grateful to an anonymous referee whose suggestions have improved the paper
considerably. Any errors in this paper are, of course, entirely my own.
those having enough wealth are in the position to start a profitable business.'
According to this interpretation, there is a causal link between the endowment of
wealth and the entrepreneurial choice. However, an inverse causation can also be
hypothesized: business families own more wealth because they save more. There
are several factors, which may account for this. For instance, the presence of
liquidity constraints may induce those families with higher entrepreneurial ability
to accumulate the capital required to start a business. Another reason may stem
from the fact that agents are risk averse and, in order to face the entrepreneurial
risk, they accumulate more assets. Finally, the higher saving rate of business
families may be a consequence of intermediation costs that make external financ-
ing more expensive; thereby implying that entrepreneurs with a lower level of
wealth have a higher marginal return from saving. In other words, while it is
possible that the presence of liquidity constraints has the effect of selecting entrep-
reneurs among richer families, it is equally plausible that these families have
higher levels of wealth relative to income, because their members have more
incentives to save.
If business families were to own more wealth only because they were selected
among richer families, then entrepreneurship would not have any implications for
wealth inequality, being the inequality in asset holdings that discriminates
between workers and entrepreneurs and not vice-versa. In other words, entrepren-
eurship has positive implications for wealth concentration-in addition to the
concentration induced by higher asset holdings as a consequence of higher busi-
ness incomes-only if entrepreneurs accumulate more wealth than workers. In
order to exploit this possibility,
I
estimate a dynamic accumulation equation on
a sample of U.S. families, and I test the hypothesis that workers and entreprene-
urs have different saving behaviors. The result of this test supports the hypothesis
that enterprising households have a higher targeted wealth-income ratio than
workers, and therefore, higher saving rates.
By looking at the "accumulation" of wealth, rather than at its "holding,"
the analysis is shifted from the static aspects of the wealth distribution to its
dynamics, namely, the movement of the agents inside the distribution or socio-
economic mobility. There are several empirical studies analyzing income and
earnings mobility. Some studies document intergenerational mobility, such as
Behrman and Taubman (1990), Solon (1992) and Zimmerman (1992); while
others concentrate on the mobility of the same individual or family, notably Dun-
can and Morgan (1984), Sawhill and Condon (1992) and Hungerford (1993).
These studies, however, do not distinguish among different types of individuals
or households and do not extend the analysis to the study of mobility within
wealth classes, other than income and earnings. In contrast, this study is primarily
interested in analyzing the mobility properties experienced by different economic
agents within one generation-namely enterprising households as compared to
other households-where the position in the social ladder is identified with
wealth.
'see, for example, Evans and Jovanovic (1989), Evans and Leighton (1989) and Holtz-Eakin
et
al.
(1994). Another interpretation is based on the selection mechanism through which only successful
entrepreneurs survive and we only observe the upper tail of the distribution.
In the data analysis of Section
2,
I
show that enterprising households experi-
ence greater upward-mobility than other households, that is, they face greater
probabilities of moving to a higher wealth class. Moreover, this upward mobility
is not merely a consequence of their higher incomes, since enterprising households
also experience greater upward mobility in the ratio of wealth to income as well.
This wealth mobility property is consistent with the observation of higher asset
holdings of entrepreneurs in the sense that entrepreneurs own more wealth
because they tend to move to higher positions in the distribution of wealth. At
the same time, the higher upward mobility of entrepreneurs can be interpreted as
evidence of the hypothesis that their higher asset holdings is not only a conse-
quence of borrowing constraints that select entrepreneurs among richer
families-
as pointed out by Evans and Jovanovic (1989)-but also a consequence of their
higher rates of savings.
The different saving patterns of workers and entrepreneurs generate higher
asset holdings in the hands of the latter and, as a result, a higher concentration
in the whole distribution of wealth. A relevant factor that determines the import-
ance of this mechanism in generating wealth concentration, is the persistence
and turnover of households in the business group. Accepting the hypothesis that
entrepreneurs save more, the amount of wealth accumulated depends on the busi-
ness duration, that is the time spent owning a business. Therefore, in Section 4, I
analyze the persistence and turnover of households in the business group and I
show that despite the high exit rates from entrepreneurship, the turnover rate in
the business group is low and a limited percentage of households tend to alternate
in the position of entrepreneur. The finding is confirmed by the estimation of a
probit model for entrance and exit to and from entrepreneurship. This low turn-
over allows a restricted group of households to accumulate consistent amounts
of wealth (due to their higher saving rates) which in turn generates higher concen-
tration of wealth.
The main source of data comes from the Panel Study of Income Dynamics
(henceforth PSID) which is a national survey conducted annually on a sample of
US.
families since 1968. The original sample included 4,800 families. Over time
the sample composition has changed due to the addition of new family units,
descending from the previous ones, and the removal of others. Although the
survey was taken annually, the main variable of interest for this study-family
wealth-is available for only a few years and the main analysis is based upon the
1984 and 1989 wealth Accordingly, the sample analyzed in this and the
next sections consists of families interviewed in all years from 1984 through 1990
and headed by the same
per~on.~
'"~arnil~ wealth" is defined as the sum of net worths of all family members and results from the
aggregation of the following components: house (main home), other real estate, vehicles, farms and
businesses, stocks, cash accounts and other assets.
3~lthough wealth data for the 1994 is also available, the paper does not extend the analysis to
the 1994 year because other variables which are needed for the analysis are not currently available.
4~he requirement that the family is headed by the same person is a way of identifying the same
family over time, and to link single years data. This link is crucial for the analysis conducted in
Sections
2-4.
The first step of this analysis is to identify business families or entrepreneurs.
I adopt two criteria. According to the first criterion, entrepreneurs are families
that own a business or have a financial interest in some business enterprise and
workers are identified as all other families.' In the second criterion, entrepreneurs
are identified as families in which the head is self-employed, in his or her main
job, while workers are identified as families in which the head is a dependent
~orker.~ Implicit in the second definition of entrepreneurs is the exclusion from
the analysis of those families in which the head is not an active worker. Hence-
forth, I will call "business owners" enterprising families identified using the first
definition of entrepreneurs, and "self-employed" enterprising families identified
using the second definition of entrepreneurs.
Figure
1
reports the fraction of business families in different wealth classes,
and for the two definitions of entrepreneurs, with each class including five percent
of all families. Given the similarity of the 1984 and 1989 data, this graph and the
main analysis of the remaining part of this section are based on the average of
these two years. Figure
1
shows that business families tend to be concentrated in
the higher wealth classes and more than half of the families located in the top
class are business families.
I
Business owners
0.5
Self-employed
1
2
3
4
5 6
7
a
9 10 11 la 13
14
15 16 17 la 19
ao
Wealth Class (Each
=
5%)
Figure 1. Fraction of Business Families in Each Wealth Class as Average of 1984 and 1989 PSID
Data
The fact that business families own more wealth than worker families would
not be of particular interest if business families also earn more income (in pro-
portion to the ownership of wealth) and a better evaluation requires the analysis
of the joint distribution of income and wealth among these two categories of
families. Based on this consideration, Figure
2
reports the average per-family
wealth of worker and business families located in each income decile, as a fraction
5~he identification of enterprising families is based on the PSID variable "Whether Business"
which is based on the following interview question:
"Did you (Head) or anyone else in the family own
a business at any time during the previous year or have a financial interest in any business enterprise?"
6~he classification is based on the following PSID interview question:
"In your main job, are you
(Head) self-employed or do you work for someone else?"
0
Workers
.
Business
owners
Figure
2.
Average Per-Family Wealth of Workers and Business Families in Each Income Decile as
Average of 1984 and 1989 PSID Data. The Top Graph Adopts the First Definition of Entrepreneurs
while the Bottom Graph Adopts the Second Definition of
Eutrepreneurs.
of total per-family wealth: the top graph for the first definition of entrepreneurs
and the bottom graph for the second definition of
entrepreneur^.^
The decile
thresholds are determined with respect to the total sample, and therefore, worker
and business families located in the same income decile dispose, approximately
and with the exception of the first and last decile, of the same income. More
detailed information is provided by Tables
9
and
10
in the Appendix.
Figure
2
shows that business families own on average higher levels of wealth,
relative to their incomes, than worker families. If we consider the total sample of
business and worker families, the ratio of wealth to income is more than twice as
large as it is for business families. In terms of total distribution, we observe that
in
1989,
and for the first definition of entrepreneurs,
14.9
percent of all families
'"Family income" is defined as the sum of incomes coming from all sources plus transfers of all
family members.
