Content uploaded by Walter W. Powell
Author content
All content in this area was uploaded by Walter W. Powell on Jul 23, 2016
Content may be subject to copyright.
Cultivating an Institutional Ecology of Organizations: Comment on Hannan, Carroll,
Dundon, and Torres
Author(s): Joel A. C. Baum and Walter W. Powell
Source:
American Sociological Review
, Vol. 60, No. 4 (Aug., 1995), pp. 529-538
Published by: American Sociological Association
Stable URL: http://www.jstor.org/stable/2096292
Accessed: 19-07-2016 21:00 UTC
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://about.jstor.org/terms
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted
digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about
JSTOR, please contact support@jstor.org.
American Sociological Association, Sage Publications, Inc.
are collaborating with JSTOR to digitize,
preserve and extend access to
American Sociological Review
This content downloaded from 171.64.184.225 on Tue, 19 Jul 2016 21:00:22 UTC
All use subject to http://about.jstor.org/terms
CULTIVATING AN
INSTITUTIONAL ECOLOGY OF ORGANIZATIONS:
COMMENT ON HANNAN, CARROLL, DUNDON, AND TORRES*
Joel A. C. Baum
University of Toronto
H annan, Carroll, Dundon, and Torres's
article on the European automobile
industry, "Organizational Evolution in a
Multinational Context" (1995, henceforward
Hannan et al.) is the most recent example of
their concerted efforts to develop a density-
dependent model of organizational evolution.
Our purpose is not to take issue with their
paper per se. Indeed, we regard incorporation
of geographic heterogeneity, and spatial and
temporal heterogeneity more broadly, as a
positive step. But we take issue with their
conceptualization of legitimacy and question
the plausibility of the underlying social pro-
cesses implied by the density-dependence
model.
As in their earlier work, Hannan et al. as-
sert that "no practical alternative to an indi-
rect approach has yet been advanced for
studying legitimation as (taken-for-granted-
ness) for long-lived populations of organiza-
tions" (p. 51 1). Their conclusion is based on
three premises: (1) Cognitive legitimacy, as
measured by organizational density, is the
most relevant facet of institutionalization; (2)
organizational density counts are comparable
across time and populations, while other
measures are post hoc and period- or popula-
tion-specific; and (3) legitimation is not a
variable to be measured, but a process that
relates organizational density to founding
and failure.
We dispute these premises. Density-depen-
dence arguments, we assert, confine thinking
about legitimation to a narrow range of insti-
tutional phenomena. Our goal is to broaden
Walter W. Powell
University of Arizona
and reorient analyses of the legitimation pro-
cess. We present alternative measurement
strategies and examine the robustness of re-
cent findings. We illustrate how ignoring the
basic historical properties of organizational
forms undermines the ostensible temporal
comparability of organizational density
counts as well as findings purported to sup-
port the model itself. We conclude that the
density-as-process view of legitimation is
difficult to maintain in light of recent non-
conforming findings.
CONCEPTUAL STATUS OF DENSITY:
COGNITIVE AND SOCIOPOLITICAL
LEGITIMACY
The new institutionalists highlight several
factors that contribute to the legitimacy of an
organizational form or practice. Zucker
(1977) treats institutionalization as a process,
emphasizing that legitimacy is a cognitive
phenomenon reflected in taken-for-granted
assumptions. Meyer and Rowan (1977) and
DiMaggio and Powell (1983) stress that le-
gitimacy is embedded in relational networks
and normative codes of conduct. Thus, they
view institutionalization as a process through
which certain activities or forms come to be
regarded as obligatory, and as a state in
which widely shared norms and values are
buttressed by cultural, professional, and po-
litical expectations or even mandated by law.
Drawing on this literature, organizational
ecologists suggest a distinction between cog-
nitive and sociopolitical legitimacy (Aldrich
and Fiol 1994).
Density-dependence theory focuses on
cognitive legitimacy: An organizational
form is legitimate "when there is little ques-
tion in the minds of actors that it serves as
the natural way to effect some kind of col-
lective action" (Hannan and Carroll
1992:34). Hannan et al.'s view sits in strong
contrast to the sociopolitical approach,
*Direct correspondence to Joel A. C. Baum,
Division of Management and Economics, Univer-
sity of Toronto, Scarborough Campus, 1265 Mili-
tary Trail, Scarborough, Ontario, MIC 1A4
(Internet: BAUM@LAKE.SCAR.UTORONTO.CA). For
helpful discussions and comments we thank
Howard Aldrich, Elisabeth Clemens, Paul Di-
Maggio, Ken Koput, Anne Miner, Christine
Oliver, James Ranger-Moore, and Huggy Rao.
American Sociological Review, 1995, Vol. 60 (August:529-538) 529
This content downloaded from 171.64.184.225 on Tue, 19 Jul 2016 21:00:22 UTC
All use subject to http://about.jstor.org/terms
530 AMERICAN SOCIOLOGICAL REVIEW
which emphasizes that embeddedness in re-
lational and normative contexts influences
an organizational form's legitimacy by sig-
naling its conformity to institutional expec-
tations. Density-dependence arguments as-
sume that sociopolitical factors are largely
derivative of population growth. Institution-
alists view these two facets of legitimacy as
complementary and fundamentally interre-
lated. Although cognitive legitimacy can be
achieved without sociopolitical approval,
the latter is considered to be a vital source
of, or impediment to, the former. Indeed,
because contemporary organizational forms
rarely operate in isolation from the state, the
professions, and broader societal influences,
sociopolitical legitimacy cannot be ignored.
Crudely put, without roads and people eager
to drive you cannot develop an automobile
industry. This point is aptly illustrated in
Sloan's (1963) account of the transforma-
tion of the U.S. automobile industry into a
mass market business. Sloan, who led Gen-
eral Motors in the 1920s and 1930s, argued
that the ascendancy of the automobile was
due to the spread and improvement of roads,
the novelty and fashion of the annual model,
and the ability of manufacturers to persuade
consumers to trade in used cars and buy new
ones on installment.
The exclusive emphasis on one facet of le-
gitimation misses its multidimensional na-
ture. We need to measure other aspects of le-
gitimation and examine how diverse social
processes combine with organizational den-
sity to contribute to the legitimacy of organi-
zational forms. After all, we have no direct
evidence that taken-for-grantedness is central
to explaining population dynamics; socio-
political legitimacy can be inferred from the
effects of density on vital rates with equal
ease. In making this point we are not sug-
gesting that legitimacy must be measured di-
rectly, but, rather, that it should be gauged
differently from organizational density using-
greater measurement precision and contex-
tual realism.' Theoretical models must bal-
ance competing pursuits of generality, real-
ism, and precision (Levins 1966). Density-
dependence arguments emphasize only gen-
erality, however. We concur with Singh
(1993) that progress in the study of the le-
gitimation of organizational forms can be en-
hanced by a move toward greater precision
and realism, even at the expense of some
generality. Although we think a similar point
underlies Hannan et al.'s concern with differ-
ent observables, replacing country-level den-
sity with European density does little to en-
hance either precision or realism, and, argu-
ably, may diminish both.
Hannan et al.'s interest in respecifying
density-dependent legitimation at a higher
level stems from their idea that cultural im-
ages flow more freely across social system
boundaries than do material resources. They
conclude that competitive environments are
country specific, while institutional environ-
ments are spatially diffuse.2 This is the fun-
damental sociological insight of their paper,
and their findings tend to support it.3 How-
ever, in their reliance on organizational den-
sity, they ignore three measurement opportu-
nities suggested by their own discussion of
legitimation.
First, as Hannan et al. point out (p. 512),
the print media are a basic source of the dif-
fusion of information about organizational
forms. Detailed archives of media coverage
exist for many industries (including the au-
tomobile industry), and content analyses of
these public records offer a potentially pow-
erful technique for operationalizing legiti-
mation. One advantage of this approach is
that the media differentially select what to
communicate about the organizational world
(due, for example, to variation in socio-
political legitimacy or network centrality of
organizations). Content analysis is widely
used in social movement research (Tarrow
I Of course, there are also sound methodologi-
cal reasons for multiple measures (e.g., Campbell
and Fiske 1959).
2We do not think sociopolitical legitimacy will
typically unfold in broader spatial contexts.
Rather, spatial boundaries of sociopolitical pro-
cesses will likely coincide with the jurisdictions
of actors in the relational and normative contexts
within which a population is embedded.
3 Hannan et al. do not actually test their hypoth-
esis that national boundaries constrain competi-
tion. Moreover, while the coefficients in their
Table 4 are generally supportive of their argu-
ments, in Table 3, where country-level density
and European density are estimated together, the
theoretically predicted pattern is found for only
two of the five countries (France and Italy).
This content downloaded from 171.64.184.225 on Tue, 19 Jul 2016 21:00:22 UTC
All use subject to http://about.jstor.org/terms
CULTIVATING AN INSTITUTIONAL ECOLOGY OF ORGANIZATIONS 531
1989; Tilly 1993); indeed, current state-of-
the-art work is done from an ecological per-
spective (Olzak 1992). Hybels (1994) dem-
onstrates the feasibility of media-based le-
gitimacy measures in his analysis of found-
ings of U.S. biotechnology firms. We think
rigorous media-based measures promise
high comparability across populations and
temporal comparability within a given
population.
Second, as Hannan et al. note, "Nothing ri-
valed the heavily publicized road races . . .
in spreading the image of the automobile" (p.
512). These races offer a unique opportunity
to examine a key source of cognitive legiti-
macy. In his work on the early U.S. automo-
bile industry, Rao (1994) argues that cumu-
lative victories in certification contests en-
hanced the reputations of manufacturers in
the eyes of risk-averse consumers and finan-
ciers. Moreover, by increasing opportunities
to disseminate knowledge about the industry,
these races established the cognitive legiti-
macy of the product and its producers. As
Rao's study shows, winning these races im-
proved the survival chances of individual
manufacturers, and in addition, the cumula-
tive prevalence of these contests lowered the
aggregate failure rate. Hannan et al. might
object that these races are an idiosyncratic
feature of the automobile industry and thus
are not generalizable. We think otherwise.
In many industries, special purpose orga-
nizations institute certification contests to
evaluate products or firms and to rank par-
ticipants according to preset criteria. For ex-
ample, J. D. Powers ranks cars on predefined
performance criteria, thus shaping the public
image of their manufacturers. Moody's rates
insurance companies, classifying them on the
basis of their viability. Consumer Reports
ranks rival products in numerous product cat-
egories. Forbes, Business Week, and Money
rank mutual funds based on their perfor-
mance. Certification contests are a common
test of products and producers, and their per-
vasiveness is reflected in the current business
practice of benchmarking. More broadly, a
wide range of readily observable credent-
ialing activities signal product reliability and
heighten the sociopolitical legitimacy of or-
ganizational forms, while contributing to the
cognitive legitimacy of organizations by dif-
fusing knowledge about them. Accreditation
frequently enables organizational forms to
create a desirable image and garner re-
sources, as does occupational licensing. Cer-
tification by established auditors and invest-
ment banks has similar consequences (Beatty
and Ritter 1986). Charitable registration may
have parallel effects for nonprofit organiza-
tions (Singh, Tucker, and House 1986). Sta-
tus-based competition in a variety of fields
yields both material and symbolic rankings,
and both serve to enhance organizational ac-
countability (Podolny 1993). In high-tech in-
dustries, patents play a similar role, estab-
lishing claims to intellectual property and
signaling competence to the investment and
scientific communities (Powell and Brantley
1992).
Third, Hannan et al. describe how "France
quickly . . . became the center of the [auto-
mobile] industry in terms of technical devel-
opment, scale of production, and number of
firms" (p. 516). Thus, the early information
that shaped legitimation of the automobile
industry most likely diffused from France to
other countries. In fact, early on, the French
did exert a pervasive influence: Their heavy
designs eclipsed lightweight buggy designs
and were copied by non-French producers
(Thomas 1965). Yet, France plays no special
role in Hannan et al.'s analysis. French pro-
ducers are combined with all others in Euro-
pean density counts. And, contrary to the im-
plied spatial diffusion process, their country-
level density coefficients imply that around
the turn of the century legitimation effects
were outweighed by within-country compe-
tition in each country's industry (see Hannan
et al., Table 2, p. 521). One way to examine
the macro-diffusion process with greater pre-
cision would be to analyze the effects of or-
ganizational density in France on the found-
ing rate in other countries. However, France
was influential not only in terms of its num-
ber of firms, but because of the character of
French automobiles. Thus, connecting the
diffusion process with French organizational
density is less precise and realistic than mea-
suring the density of French designs, defined
as the number of French and non-French pro-
ducers using French automobile designs. In-
deed, one is tempted to reverse the causal or-
der in this (and other) case(s): Underlying
technologies are reflected in the evolution of
an industry, and technological change (e.g.,
This content downloaded from 171.64.184.225 on Tue, 19 Jul 2016 21:00:22 UTC
All use subject to http://about.jstor.org/terms
532 AMERICAN SOCIOLOGICAL REVIEW
the emergence of a dominant product design)
is a critical factor that influences industrial
dynamics.
OPERATIONAL STATUS OF DENSITY:
AHISTORICISM AND
COMPARABILITY
A primary rationale for studying organiza-
tional density is that data are available and
purportedly comparable, but we fail to see
how data convenience ensures comparability.
In the case of Hannan et al., comparability is
marred by inattention to some basic histori-
cal properties of organizational forms. We
first develop this point and then show how
this inattention undermines other recent den-
sity dependence studies as well.
Ahistoricism poses two problems. The
first stems from inadequate specification of
organizational forms. The evolutionary tra-
jectories of diverse long-lived organiza-
tional forms appear to follow a common
path (Hannan and Carroll 1992). The num-
ber of organizations grows slowly initially,
then increases rapidly to a peak. Once the
peak is reached, the number often declines
as a few large organizations begin to domi-
nate many smaller ones. This pattern is of
substantive interest because it has implica-
tions for the distribution of power in organi-
zational populations. Populations of news-
paper publishers, labor unions, breweries,
and banks follow this basic pattern (Hannan
and Carroll 1992), as do European automo-
bile manufacturers, which evolved from
highly fragmented national industries com-
prised of small craft producers, to a concen-
trated international industry dominated by a
small number of mass producers. Indeed,
early on, automobile manufacturers didn't
actually exist-in 1894 the world's leading
car company, Panhard et Levassor, was a
machine-tool company (Womack, Jones,
and Roos 1990:21). Yet, because the organi-
zational form is defined simply as "firms
that make autos for sale," early craft produc-
ers are treated as equal to Citroen, Fiat,
Mercedes, Morris, Renault, and Volks-
wagen, which by the mid-1950s were mass-
producing cars at a scale comparable to ma-
jor U.S. facilities. Clearly the shift from
craft to mass production changed the face of
the auto industry, and with this change, we
argue, the meaning of organizational density
was transformed as well.
A second, closely related problem is that
density-dependence theory is commonly
tested in long-lived populations, like Euro-
pean automobile manufacturers, that have
become concentrated over time. Since den-
sity-dependence arguments assume that all
members of a population contribute to and
experience competition (and legitimation)
equally, no single organization or small
group of organizations can dominate compe-
tition. Consequently, density dependence can
explain the shape of the growth trajectory of
organizational populations to an equilibrium
size, but cannot explain their subsequent
concentration (Hannan and Carroll 1992).
When tested in populations that have become
concentrated, interpretations of density-de-
pendent legitimacy are obscured in two ways
(Baum forthcoming). The meaning of low
organizational density varies greatly between
the early and late periods of a population's
history. In early low-density conditions, no
single organization or group dominates; new
entrants are typically similar to incumbents.
By comparison, in late low-density condi-
tions, organizations tend to differ consider-
ably in size and strategy, and one or a few
large organizations often dominate the mar-
ket. Second, when density is low early and
late in a population's history, the interpreta-
tion of the linear density term is unclear: Al-
though early low density has a specific role
in density-dependence theory (i.e., legitima-
tion), late low density remains unexplained.
Early and late low organizational density
conditions, however, seem likely to have par-
allel effects on vital rates (although for very
different reasons) that are not distinguished
by empirical estimates: Large organizations
that control substantial market segments are
unlikely to fail, and increasing concentration
may create opportunities for new, specialized
entrants (Carroll 1985). Thus a supportive
linear density term may indicate late market
power and resource partitioning, not early or-
ganizational legitimation (Baum forthcom-
ing).
The foregoing arguments lead us to ques-
tion tests of density dependence in popula-
tions that have experienced both early and
late low-density conditions. We examine
density dependence studies published since
This content downloaded from 171.64.184.225 on Tue, 19 Jul 2016 21:00:22 UTC
All use subject to http://about.jstor.org/terms
CULTIVATING AN INSTITUTIONAL ECOLOGY OF ORGANIZATIONS 533
1990 to assess the implications of including
late low-density periods.4 Founding studies
(including Hannan et al.) support the theory's
predictions in 23 of 31 (74 percent) samples.
Among failure studies, support is lower (12
of 22 samples, or 55 percent).5 Notably, sup-
port for density dependence is stronger in
samples that have passed their peak density,
jumping to 82 percent (14 of 17 samples) for
foundings and 64 percent (7 of 11 samples)
for failures. Thus, the findings of numerous
studies supporting density-dependence
theory may be undermined by the incorpora-
tion of information on the decline of the
population, which density-dependence
theory is not designed to explain. This is es-
pecially true for studies by Hannan and his
colleagues, many of which test the density-
dependence model in organizational popula-
tions that have evolved (often well) beyond
their peak densities.
EMPIRICAL STATUS OF DENSITY:
PROXY VERSUS PROCESS
Hannan et al. defend their continued exclu-
sive focus on organizational density with a
series of interrelated arguments. Their first
claim is that cognitive legitimacy defies di-
rect measurement, and consequently, it
makes sense to study variables that can be
easily observed and compared across popu-
lations and over long periods of time, such
as density and vital rates. This indirect mea-
surement approach has been criticized be-
cause confirming findings cannot be inter-
preted precisely: Estimates reveal little about
the theoretical explanations designed to ac-
count for them (Singh 1993). Critics question
the legitimacy interpretation of density ef-
fects, suggesting that legitimation is invoked
ex-post facto and that density estimates are
proxies for a wide range of other possible ef-
fects (Delacroix and Rao 1994; Petersen and
Koput 1991; Zucker 1989).
These critiques appear to have led to the
density-as-process argument, in which legiti-
mation is no longer a variable to be mea-
sured, but a process that relates density to
founding and failure rates. Thus, Hannan and
Carroll (1992) claim that "growth in density
controls . . [legitimation] processes-it
does not reflect them" (p. 69). These com-
peting proxy and process views suggest dif-
ferent effects of adding covariates. Accord-
ing to Hannan and Carroll (1992:70-71), if
density is an indirect indicator of legitima-
tion, measuring legitimation more directly
would dampen the first-order effects of den-
sity or lead them to disappear altogether. But
from the density-as-process view, the inclu-
sion of such covariates implies a sharpening
and strengthening of the legitimation effects
of density. Ultimately, however, Hannan and
Carroll conclude that multicollinearity would
make it impossible to test their expectations;
thus they cut off debate.
Although Hannan et al. now endorse a
strong distinction between taken-for-granted-
ness and external endorsements (note 3, p.
511), in earlier writings (Hannan and Carroll
1992:41) they clearly identify sponsorship by
powerful actors and organizations as an im-
portant mechanism through which organiza-
tional forms can gain acceptance and cogni-
tive legitimacy. Baum and Oliver (1992) ex-
plain the legitimation of an organizational
form in terms of such endorsements and pro-
4 We do not include in our summary Hannan
and Carroll's (1992) multiple subanalyses of U.S.
brewers at regional, state, and city levels. Since
many nonlocal density specifications also support
density dependence predictions, these analyses
appear undermined by misspecification of the
spatial boundaries. We include one supportive
unpublished study of U.S. automobile manufac-
turers by Rao (1992) for its comparative rel-
evance. Aside from Rao, we do not include un-
published papers in Table 1.
5 In Table 1, we can also examine Hannan and
Carroll's (1992) claim that nonsupportive studies
fail to incorporate data on the early years of a
population's history (i.e., are left-truncated).
Founding studies support the density-dependence
predictions equally for left-truncated (three of
four samples, or 75 percent) and complete
samples (20 of 27 samples, or 74 percent).
Among failure studies, however, support is higher
for complete (10 of 16 samples, or 63 percent)
than for left-truncated samples (2 of 6 samples,
or 33 percent). But, in the unsupportive, left-trun-
cated studies, the coefficients do not consistently
yield the purely competitive effect implied by
Hannan and Carroll's argument. So although
there is some evidence that support for density-
dependent organizational failure is stronger in
studies of untruncated population histories, the
results of unsupportive truncated studies do not
conform to their expectation.
This content downloaded from 171.64.184.225 on Tue, 19 Jul 2016 21:00:22 UTC
All use subject to http://about.jstor.org/terms
534 AMERICAN SOCIOLOGICAL REVIEW
Table 1. Summary of Studies Testing the Density-Dependence Model, 1990-1994
Sign of Coefficient
Population Density (Density)2 Source
Organizational Foundings
U.S. labor unions, 1836-1985b + - Hannan and Carroll (1992);
U.S. craft labor unions, 1836-1985b + - Ranger-Moore, Banaszak-Holl,
U.S. industrial labor unions, 1853-1985 n.s. - and Hannan (1991)
U.S. breweries, 1633- 1988b c + -
San Francisco newspapers, 1800-1975b + -
Argentina newspapers, 1800-1900 + -
Ireland newspapers, 1800-1975b n.s. n.s.
Manhattan banks, 1791-1980b +
Manhattan commercial Banks, 1792-1980b +
Manhattan savings banks, 1820-1980b + n.s.
U.S. life insurance companies, 1759-1937 +
U.S. stock life insurance companies, 1787-1937 +
U.S. mutual life insurance companies, 1759-1937 +
Pennsylvania telephone companies, 1879-1934 - + Barnett (1990; Barnett and
Amburgey (1990)
U.S. semiconductor firms, 1947-1984 + + Hannan and Freeman (1989);
Freeman (1990)
Toronto voluntary social service organizations, n.s. n.s. Singh, Tucker, and Meinhard
1970-1982a (1991)
Toronto daycare centers, 1971-1989a.d + - Baum and Oliver (1992); Baum
and Singh (1994a, 1994b)
U.S. automobile industry, 1893-1915 + - Rao (1992)
New York state life insurance companies, n.s. + Budros (1993, 1994)
1842-1904
German breweries, 1861-1988b + n.s. Carroll, Preisendoerfer, Swamin-
athan, and Wiedenmayer (1993)
U.S. trade associations, 1901_1990ab + - Aldrich, Zimmer, Staber, and
Beggs (1994)
Niagara Falls hotels, 1894-199 ab + - Ingram and Inman (1994)
U.S. biotechnology firms, 1971-1992 + - Hybels, Ryan, and Barley (1994)
Manhattan analog fax transmission companies, + - Baum, Korn, and Kotha
1965-1992b (1995)
Manhattan digital fax transmission companies, + -
1981-1992
Organizational Failures
U.S. labor unions, 1836-1985b - n.s. Hannan and Carroll (1992);
U.S. craft labor unions, 1836-1985b - + Ranger-Moore, Banazak-Holl,
U.S. industrial labor unions, 1853-1985 n.s. n.s. and Hannan (1991)
U.S. breweries, 1633- 1988b.c - +
San Francisco newspapers, 1800-1975b - +
Argentina newspapers, 1800-1900 - +
Ireland newspapers, 1800-1975b - +
Manhattan banks, 1791-1980b - +
U.S. life insurance companies, 1759-1900 n.s. n.s.
Pennsylvania telephone companies, 1879-1934 + n.s. Barnett (1990); Barnett and
Southeast Iowa telephone companies, 1900-1929a + n.s. Amburgey (1990)
U.S. semiconductor firms, 1947-1984 - + Hannan and Freeman (1989);
Freeman (1990)
(Table I continued on next page)
This content downloaded from 171.64.184.225 on Tue, 19 Jul 2016 21:00:22 UTC
All use subject to http://about.jstor.org/terms
CULTIVATING AN INSTITUTIONAL ECOLOGY OF ORGANIZATIONS 535
(Table I continued)
Sign of Coefficient
Population Density (Density)2 Source
Toronto voluntary social service organizations, + - Singh, Tucker, and Meinhard
1970-1982a (1991)
Bavarian breweries, 1900-1981ab n.s. - Swaminathan and Wiedemayer
(1991)
Toronto daycare centers, 1971-1989ad - + Baum and Oliver (1992); Baum
and Singh (1994a, 1994b)
Manhattan hotels, 1898-1990ab - n.s. Baum and Mezias (1992)
German breweries, 1861-1988b - + Carroll, Preisendoerfer, Swamina-
than, and Wiedenmayer (1993)
U.S. trade associations, 1901-1990a b - + Aldrich, Zimmer, Staber, and
Beggs (1994)
U.S. credit unions, 1980-1989a n.s. n.s. Amburgey, Dacin, and Kelly
(1994)
New York City credit unions, 1914-1990b - + Barron, West, and Hannan (1994)
U.S. automobile industry, 1895-1912 + n.s. Rao (1994)
Manhattan analog fax transmission companies, + n.s. Baum, Korn, and Kotha
1965-1992 b (1995)
Manhattan digital fax transmission companies, - +
1981-1992
a Sample design left-truncated (i.e., some organizations fail before observation period begins).
b Sample has declined in numbers from its peak by 25 percent or more.
c Support for founding and failure when sample left-truncated to 1800, 1878, and 1891.
d Support for founding when sample further left-truncated to 1978.
pose that, in contexts where relations with
community and government are dense, legiti-
mation may be accounted for by a
population's relational embeddedness. They
model the embeddedness of a population in
its institutional environment by measuring
relational density (the number of relation-
ships between the members of a population
and community organizations and govern-
ment agencies). In addition to greater preci-
sion and contextual realism, relational den-
sity has the potential advantages of compara-
bility across organizational contexts and tem-
poral comparability within contexts, when the
nature and form of relations remain stable
over a population's history. While initial esti-
mates in their study of daycare centers sup-
port curvilinear density dependence predic-
tions for both founding and failure rates, the
"legitimating" effects of initial increases in
organizational density disappeared after the
inclusion of relational density and the rela-
tionship between organizational density and
founding and failure rates became purely
competitive. Thus, Baum and Oliver (1992)
test the proxy-versus-process prediction, and
their results support the proxy view. Their
findings have been replicated in Hybels et
al.'s (1994) study of the founding of U.S. bio-
technology firms in which they use strategic
alliances to measure embeddedness in rela-
tional and institutional contexts. These stud-
ies suggest that the initial density-as-proxy
formulation of legitimacy was more accurate
and, in addition, that organizational density
may be a proxy for relational as well as (or
instead of) cognitive legitimacy.
TOWARD AN INSTITUTIONAL
ECOLOGY OF ORGANIZATIONS
Legitimation is a core sociological concept;
nevertheless, its dual nature, as both process
and outcome, has posed thorny measurement
problems. Given these difficulties, density
dependence has considerable appeal. It is
This content downloaded from 171.64.184.225 on Tue, 19 Jul 2016 21:00:22 UTC
All use subject to http://about.jstor.org/terms
536 AMERICAN SOCIOLOGICAL REVIEW
relatively simple to count the number of or-
ganizations in a population and to assume that
growth in numbers captures "legitimation."
We urge researchers studying organizations
to resist this temptation to simply count and
to explore new dimensions and measures of
legitimation. Contrary to Hannan et al.'s
claims, we think there are several promising
nondensity-based alternatives to studying the
legitimation of organizational forms that are
both fine-grained and generalizable.
Organizational ecologists have made con-
siderable progress in demonstrating that cul-
tural understandings play a basic role in the
ecological dynamics of organizational popu-
lations. We applaud these efforts, but contend
that the evolutionary dynamics of organiza-
tional populations transcend density depen-
dence to include sociopolitical forces that
provide vital support for organizational de-
velopment. We stress as well that the devel-
opment of population-wide norms and prac-
tices and support from key institutional ac-
tors is, in part, the product of competitive
struggles, and thus is vulnerable to resource
constraints. There is much that ecological
and institutional arguments have to offer one
another. Clearly, for ecological theory to
move from demographics to truly evolution-
ary analyses of organizations, it must use
better and more robust tools to incorporate
the historical processes that link organiza-
tions through time (Baum and Singh 1994c).
Similarly, for institutional analysis to capture
the dynamic aspects of institutionalization, it
must yield insights into the differential na-
ture of ecological interaction and replication
(Powell 1991). We suggest that future efforts,
comparable to those devoted to analyzing
density dependence, focus on constructing an
institutional ecology of organizations-a
more sophisticated theoretical understanding
of the co-evolving nature of cultural under-
standings, organizational forms, and resource
constraints.
Joel A. C. Baum is Associate Professor of Man-
agement in the Division of Management and Eco-
nomics at the University of Toronto. His research
focuses on ecological, institutional, and evolu-
tionary processes in organizational populations.
In his current work on the Manhattan hotel in-
dustry, he is studying managers' (rnis)cate-
gorization of competitors (with Theresa K. Lant),
spatial evolution (with Heather A. Haveman), and
the dynamics of chain affiliation (with Paul L.
Ingram).
Walter W. Powell is Professor of Sociology at the
University of Arizona. He is presently studying
the development of the biotechnology industry,
with particular interest in the relations between
elite universities and science-based companies.
REFERENCES
Aldrich, Howard E. and C. Marlene Fiol. 1994.
"Fools Rush In? The Institutional Context of
Industry Creation." Academy of Management
Review 19:645-70.
Aldrich, Howard E., Catherine R. Zimmer, Udo
H. Staber, and John J. Beggs. 1994. "Mini-
malism, Mutualism, and Maturity: The Evolu-
tion of the American Trade Association Popu-
lation in the 20th Century." Pp. 223-39 in Evo-
lutionary Dynamics of Organizations, edited
by J. A. C. Baum and J. V. Singh. New York:
Oxford University Press.
Amburgey, Terry L., Tina Dacin, and Dawn
Kelly. 1994 "Disruptive Selection and Popula-
tion Segmentation: Interpopulation Competi-
tion as a Segregation Process." Pp. 240-54 in
Evolutionary Dynamics of Organizations, ed-
ited by J. A. C. Baum and J. V. Singh. New
York: Oxford University Press.
Barnett, William P. 1990. "The Organizational
Ecology of a Technical System." Administra-
tive Science Quarterly 35:31-60.
Barnett, William P. and Terry L. Amburgey.
1990. "Do Larger Organizations Generate
Stronger Competition?" Pp. 78-102 in Organi-
zational Evolution: New Directions, edited by
J. V. Singh. Newbury Park, CA: Sage.
Barron, David N., Elizabeth West and Michael T.
Hannan. 1994. "A Time to Grow and a Time to
Die: Growth and Mortality of Credit Unions in
New York City, 1914-1990." American Jour-
nal of Sociology 100:381-421.
Baum, Joel A. C. Forthcoming. "The Changing
Basis of Competition in Organizational Popu-
lations: Evidence from the Manhattan Hotel In-
dustry, 1898-1990." Social Forces 74(1).
Baum, Joel A. C., Helaine J. Korn, and Suresh
Kotha. 1995. "Dominant Designs and Popula-
tion Dynamics in Telecommunications Ser-
vices: Founding and Failure of Facsimile Ser-
vice Organizations, 1965-1992." Social Sci-
ence Research 24:97-135.
Baum, Joel A. C. and Stephen J. Mezias. 1992.
"Localized Competition and Organizational
Failure in the Manhattan Hotel Industry, 1898-
1990. Administrative Science Quarterly 37:
580-604.
Baum, Joel A. C. and Christine Oliver. 1992. "In-
stitutional Embeddedness and the Dynamics of
This content downloaded from 171.64.184.225 on Tue, 19 Jul 2016 21:00:22 UTC
All use subject to http://about.jstor.org/terms
CULTIVATING AN INSTITUTIONAL ECOLOGY OF ORGANIZATIONS 537
Organizational Populations." American Socio-
logical Review 57:540-59.
Baum, Joel A. C. and Jitendra V. Singh. 1994a.
"Organizational Niches and the Dynamics of
Organizational Mortality." American Journal
of Sociology 100:346-80.
. 1994b. "Organizational Niches and the
Dynamics of Organizational Founding." Orga-
nization Science 5:483-501.
-. 1994c. "Organizational Hierarchies and
Evolutionary Processes: Some Reflections on
a Theory of Organizational Evolution." Pp. 3-
20 in Evolutionary Dynamics of Organizations,
edited by J. A. C. Baum and J. V. Singh. New
York: Oxford University Press.
Beatty R. and J. Ritter. 1986. "Investment Bank-
ing, Reputation, and the Underpricing of Ini-
tial Public Offerings." Journal of Financial
Economics 15:213-32.
Budros, Art. 1993. "The Making of an Industry:
Organizational Births in New York's Life In-
surance Industry." Social Forces 70:1013-33.
. 1994. "Analyzing Unexpected Density
Dependence Effects on Organizational Births
in New York's Life Insurance Industry." Or-
ganization Science 5:541-53.
Campbell, Donald T. and Donald W. Fiske. 1959.
"Convergent and Discriminant Validation by
the Multitrait-Multimethod Matrix." Psycho-
logical Bulletin 59:81-105.
Carroll, Glenn R. 1985. "Concentration and Spe-
cialization: Dynamics of Niche Width in Popu-
lations of Organizations." American Journal of
Sociology 90:1262-83.
Carroll, Glenn R., Peter Preisendoerfer, Anand
Swaminathan, and Gabriele Wiedenmayer.
1993. "Brewery and Braueri: The Organiza-
tional Ecology of Brewing." Organization
Studies 14:155-88.
Delacroix, Jacques and Hayagreeva Rao. 1994.
"Externalities and Ecological Theory: Unbun-
dling Density Dependence." Pp. 255-68 in
Evolutionary Dynamics of Organizations, ed-
ited by J. A. C. Baum and J. V. Singh. New
York: Oxford University Press.
DiMaggio, Paul J. and Walter W. Powell. 1983.
"The Iron Cage Revisited: Institutional Isomor-
phism and Collective Rationality in Organiza-
tional Fields." American Sociological Review
48:147-60.
Freeman, John. 1990. "Ecological Analysis of
Semiconductor Firm Mortality." Pp. 53-77 in
Organizational Evolution: New Directions, ed-
ited by J. V. Singh. Newbury Park, CA: Sage.
Hannan, Michael T. and Glenn R. Carroll. 1992.
Dynamics of Organizational Populations: Den-
sity, Competition, and Legitimation. New
York: Oxford University Press.
Hannan, Michael T., Glenn R. Carroll, Elizabeth
A. Dundon, and John C. Torres. 1995. "Orga-
nizational Evolution in a Multinational Con-
text: Entries of Automobile Manufacturers in
Belgium, Britain, France, Germany, and Italy."
American Sociological Review 60:509-528.
Hannan, Michael T. and John Freeman. 1989.
Organizational Ecology. Cambridge, MA:
Harvard University Press.
Hybels, Ralph C. 1994. "Legitimation, Popula-
tion Density, and Founding Rates: The Institu-
tionalization of Commercial Biotechnology in
the U.S., 1971-1989. Ph.D. dissertation,
Johnson Graduate School of Management,
Cornell University, Ithaca, NY.
Hybels, Ralph, C., Alan R. Ryan, and Stephen R.
Barley. 1994. "Alliances, Legitimation, and
Founding Rates in the U.S. Biotechnology
Field, 1971-1989." Paper presented at the
Academy of Management meetings, August,
Dallas, TX.
Ingram, Paul L. and Crist Inman. 1994. "The
Pragmatics of Cooperation: Rivalry, Ribaldry,
and Recourse at Niagara Falls." Pp. 213-17 in
Academy of Management Best Papers Proceed-
ings, edited by D. P. Moore. Dallas, TX: Acad-
emy of Management.
Levins, Richard. 1966. "The Strategy of Model
Building in Population Biology." American
Scientist 54:421-31.
Meyer, John W. and Brian Rowan. 1977. "Insti-
tutionalized Organizations: Formal Structure as
Myth and Ceremony." American Journal of So-
ciology 83:340-63.
Olzak, Susan. 1992. Dynamics of Ethnic Compe-
tition and Conflict. Stanford, CA: Stanford
University Press.
Petersen, Trond and Kenneth W. Koput. 1991.
"Density Dependence in Organizational Mor-
tality: Legitimacy or Unobserved Heterogene-
ity." American Sociological Review 56:399-
409.
Podolny, Joel M. 1993. "A Status-Based Model
of Market Competition." American Journal of
Sociology 98:829-72.
Powell, Walter W. 1991. "Expanding the Scope
of Institutional Analysis." Pp. 183-203 in The
New Institutionalism in Organizational Analy-
sis, edited by W. W. Powell and P. J.
DiMaggio. Chicago, IL: University of Chicago
Press.
Powell, Walter W. and Peter Brantley. 1992.
"Competitive Cooperation in Biotechnology:
Learning through Networks." Pp. 366-94 in
Networks and Organizations, edited by N.
Noria and R. Eccles. Boston, MA: Harvard
Business School Press.
Ranger-Moore, James, Jane J. Banaszak-Holl,
and Michael T. Hannan. 1991. 'Density De-
pendent Dynamics in Regulated Industries:
Founding Rates of Banks and Life Insurance
Companies." Administrative Science Quarterly
This content downloaded from 171.64.184.225 on Tue, 19 Jul 2016 21:00:22 UTC
All use subject to http://about.jstor.org/terms
538 AMERICAN SOCIOLOGICAL REVIEW
36:36-65.
Rao, Hayagreeva. 1992. "An Ecology of Emerg-
ing Organizations: Incorporation and Opera-
tional Startup in the American Automobile In-
dustry; 1893-1915." Goizueta Business
School, Emory University, Atlanta, GA. Un-
published manuscript.
.1994. "The Social Construction of Repu-
tation: Certification Contests, Legitimation and
the Survival of Organizations in the U.S. Au-
tomobile Industry; 1895-1912. Strategic Man-
agement Journal 15(S2):29-44.
Singh, Jitendra V. 1993. "Review Essay: Density
Dependence Theory-Current Issues, Future
Promise." American Journal of Sociology 99:
464-73.
Singh, Jitendra V., David J. Tucker, and Robert
J. House. 1986. "Organizational Legitimacy
and the Liability of Newness." Administrative
Science Quarterly 31:171-93.
Singh, Jitendra V., David J. Tucker, and Agnes
G. Meinhard. 1991. "Institutional Change and
Ecological Dynamics." Pp. 390-422 in The
New Institutionalism in Organizational Analy-
sis, edited by W. W. Powell and P. J.
DiMaggio. Chicago, IL: University of Chicago
Press.
Sloan Jr., Alfred P. 1963. My Years with General
Motors. New York: Doubleday.
Swaminathan, Anand and Gabriele Wiedenmayer.
1991. "Does the Pattern of Density Depen-
dence in Organizational Mortality Rates Vary
Across Levels of Analysis? Evidence from the
German Brewing Industry." Social Science Re-
search 20:45-73.
Tarrow, Sidney. 1989. Democracy and Disorder.
New York: Oxford University Press.
Thomas, Robert Paul. 1965. "An Analysis of the
Pattern of Growth of the Automobile Industry:
1895-1929." Ph.D. dissertation, Department of
Economics, Northwestern University, Chicago,
IL.
Tilly, Charles. 1993. European Revolutions,
1492-1992. Cambridge, MA: Blackwell.
Womack, James P., Daniel T. Jones, and Daniel
Roos. 1990 The Machine That Changed the
World. New York: Rawson Associates.
Zucker, Lynne G. 1977. "The Role of Institution-
alization in Cultural Persistence." American
Sociological Review 42:726-43.
. 1989. "Combining Institutional Theory
and Population Ecology: No Legitimacy, No
History." American Sociological Review 54:
542-45.
This content downloaded from 171.64.184.225 on Tue, 19 Jul 2016 21:00:22 UTC
All use subject to http://about.jstor.org/terms
