The Fruits Of Legitimacywhy Some New Ventures Gain More From Innovation Than Others

Description
During this detailed paper regarding the fruits of legitimacywhy some new ventures gain more from innovation than others.

58
Journal of Marketing
Vol. 72 (July 2008), 58–75
© 2008, American Marketing Association
ISSN: 0022-2429 (print), 1547-7185 (electronic)
Raghunath Singh Rao, Rajesh K. Chandy, & Jaideep C. Prabhu
The Fruits of Legitimacy: Why Some
New Ventures Gain More from
Innovation Than Others
This article examines rewards to product introduction by new ventures. The authors argue that the new ventures
that gain the most from innovation are those that adopt strategies that give them legitimacy in the eyes of
stakeholders. New ventures can gain legitimacy by creating associations with established entities; such
associations can be internal or external to the firm. The authors test these ideas by examining the stock market
gains of all products introduced between 1982 and 2002 by all public firms in the U.S. biotechnology industry. The
results show that new ventures that acquire legitimacy externally by forming alliances with established firms gain
more from their new products than new ventures that do not form such alliances. Among new ventures that do not
form alliances, those that acquire legitimacy internally by creating a history of product launches or by hiring reputed
executives or scientists gain more from their new products than those that do not. In relative terms, it pays more to
have introduced a new drug before than to have a reputed executive on the firm’s board; in turn, adding a reputed
executive pays more than adding a reputed scientist to the firm’s board. Finally, although new ventures can gain
from either external or internal legitimacy, pursuit of external legitimacy by firms that already have internal
legitimacy leads to lower rewards to innovation.
Keywords: innovation, stock market returns, legitimacy, biotechnology, alliances, entrepreneurship
Raghunath Singh Rao is Assistant Professor of Marketing, McCombs
School of Business, University of Texas at Austin (e-mail: Raghunath.
[email protected]). Rajesh K. Chandy is James D. Watkins
Chair and Professor of Marketing, Carlson School of Management, Uni-
versity of Minnesota (e-mail: [email protected]). Jaideep C. Prabhu is
Professor of Marketing and AIM Innovation Fellow, Tanaka Business
School, Imperial College London (e-mail: [email protected]). The
authors thank Kersi Antia; Sunil Kishore; Om Narasimhan; Nikos Nico-
laou; Akshay Rao; Scott Shane; K.K. Sinha; and participants at seminars
at the University of Houston, University of Minnesota, the Marketing
Science Conference at the University of Maryland, and the American
Marketing Association Winter Educators’ Conference in San Antonio for
their valuable feedback.
world are buzzing with activities that embody their hopes
and dreams. However, history suggests that only a few of
these dreams are likely to be fulfilled (Klepper 2002). Why
do some new ventures reap the rewards from their new
products while others fail to do so? This is the question we
attempt to answer in this research.
As in previous studies (Deeds, Mang, and Frandsen
2004), we define an emerging industry as one that is created
around new technologies. Recent examples of such indus-
tries include biotechnology and Web retailing. In line with
previous studies (e.g., Van de Ven, Hudson, and Schroeder
1984), we define a new venture as one whose primary busi-
ness is in an emerging industry.
Much research on the differential gains from product
introduction has studied firms in mature industries (see,
e.g., Datar and Jordon 1997; Song and Parry 1997). Prior
research has examined the impact on gains from product
introduction due to variables such as firm size (Chaney,
Devinney, and Winer 1991), resources (Sorescu, Chandy,
and Prabhu 2003), the nature of product offerings (Calan-
tone, Schmidt, and Song 1996), and industry characteristics
(Chaney, Devinney, and Winer 1991). Although these
variables are important in the context of mature industries,
it is less clear how their importance transfers to a new ven-
ture context. Within emerging industries, firms are often
similar in size and resources; they are mostly small and
have limited financial resources. Although differences
among products exist, many new products in emerging
industries are, almost by definition, all breakthroughs in
nature. As we show subsequently, variables that are tradi-
tionally examined in the literature (e.g., firm size, resources,
the nature of the product itself) play a less prominent role in
N
ew ventures play a crucial role in driving economic
growth (Schumpeter 1934). Economies with a
higher proportion of new ventures grow faster than
others (Schmitz 1989). The actions of new ventures may
even spur large, incumbent firms into action, thus accelerat-
ing the pace of technological change (Sorescu, Chandy, and
Prabhu 2003). For consumers, the creative energies of new
ventures yield a bountiful variety of new products put forth
by these firms as they maneuver for success and survival.
Indeed, the ability to introduce successful new products
is critical to the survival and growth of new ventures in
emerging industries (Dowell and Swaminathan 2000).
Product introductions generate revenue, reduce mortality,
and help predict future success (Mata and Portugal 1994).
However, the rewards to product introductions are not dis-
tributed equally across firms. Competitors with seemingly
similar products nevertheless meet wildly different fates. At
this very moment, labs (or garages) of inventors around the
The Fruits of Legitimacy / 59
separating winners from losers in the new product game in
emerging industries (see Shane 2001).
A particularly vexing challenge for firms in emerging
industries is the “liability of newness” they face (Stinch-
combe 1965). Potential stakeholders view firms in these
industries with skepticism (Hannan and Freeman 1984). For
example, venture capitalists keenly observe the actions of
new ventures and choke off funds if new products do not
show promise at the earliest stages of introduction. This
skepticism makes it difficult for such firms to gain access to
the funds they crucially need for survival and growth.
Indeed, this leads to a Catch-22 situation: To overcome
stakeholder skepticism, new ventures need successful new
products, but to have successful products, they must over-
come stakeholder skepticism in the first place. An important
way that new ventures can overcome the liability of new-
ness and increase their gains from new products is by taking
actions that provide them with legitimacy in the eyes of
stakeholders (Deeds, Mang, and Frandsen 2004; Dowling
and Pfeffer 1975). We argue that new ventures can gain
legitimacy by creating associations with more established
entities, either external or internal to the firm. In turn, this
legitimacy pays off by raising the rewards to such firms
from their product introductions.
We address the following research questions in the con-
text of new ventures:
•What are the means by which new ventures can gain
legitimacy?
•How do the different types of legitimacy affect the rewards to
new product introduction?
•What are the financial consequences of concurrently pursuing
internal and external legitimacy?
This is the first study to examine and empirically test
the effects of legitimacy on the rewards to product introduc-
tion. We argue that the impact of legitimacy may not be
straightforward. Although legitimacy can have beneficial
effects, duplication of legitimacy leads to lowered marginal
returns to product introductions by new ventures. We test
our arguments by examining the stock market gains of all
products introduced by public firms in the U.S. biotechnol-
ogy industry in its emergent phase (1982–2002).
This article makes the following contributions: First,
our study outlines several managerially controllable factors
that drive the rewards associated with new ventures’ product
introductions. By outlining key drivers of rewards to prod-
uct introduction, we provide a framework that managers of
such firms can use to manage product launch.
Second, we measure the rewards to product introduction
in terms of the value new products create for firms through
the stock market. By examining the impact of product-
related actions on stock market performance, we contribute
to the research on the marketing–finance interface. In par-
ticular, we respond to calls to adopt measures that link mar-
keting activities to firm performance and shareholder value
(Srivastava, Shervani, and Fahey 1999).
Finally, we highlight the crucial role of marketing in
entrepreneurial settings. The vast literature on entrepreneur-
ship draws on economics, sociology, and management but
rarely on marketing. Moreover, the marketing literature is
itself mostly silent on marketing’s influence on the perfor-
mance of new ventures (see Matsuno, Mentzer, and
Ozsomer 2002). This article fills this gap by emphasizing
the unique role of product innovation and legitimizing
strategies in driving the growth of new ventures.
Theory
What Is Legitimacy?
At the height of his wealth and success, the financier
Baron de Rothschild was petitioned for a loan by an
acquaintance. Reputedly, the great man replied, “I won’t
give you a loan myself, but I will walk arm-in-arm with
you across the floor of the Stock Exchange, and soon you
shall have willing lenders to spare.” (Cialdini 1989, p. 45)
Legitimacy is “a generalized perception or assumption
that the actions of an entity are desirable” (Suchman 1995,
p. 574; see also Higgins and Gulati 2006). In emerging
industries, all firms, but especially new ventures, suffer
from a lack of legitimacy in the eyes of important stake-
holders, such as venture capitalists, stock market analysts,
and consumers (DiMaggio 1988). This lack of legitimacy is
partly what creates a liability of newness for such firms
(Stuart, Hoang, and Hybels 1999). However, not all new
ventures suffer equally from this liability. Some manage to
prevail, whereas others are crushed by it. Recent research
indicates that actions, both substantive and symbolic, on the
part of new ventures can help overcome this liability (Ash-
forth and Gibbs 1990; Suchman 1995; Van de Ven, Hudson,
and Schroeder 1984; Zajac and Westphal 2004).
On the basis of this recent research, we argue that an
important means by which new ventures gain legitimacy is
by engaging in actions that convey associations with
reputed entities (see also Higgins and Gulati 2006). Such
actions affect stakeholder confidence in new ventures, in
particular in their ability to introduce products successfully.
By linking legitimacy to stakeholders’ perceptions of the
value of firms’ new products, we explain why some new
ventures gain more from their product introductions than
others.
Although there is much theorizing on the topic of legiti-
macy, empirical research is rare. Moreover, many
researchers have noted that existing studies tend to employ
a narrow, unidimensional view of legitimacy and to exam-
ine the construct using relatively crude measures (see
Deeds, Mang, and Frandsen 2004). In this study, we high-
light and empirically test the role of a variety of legitimiz-
ing actions. We also show that these legitimizing actions
may not always work together; some may actually work at
cross-purposes with others.
Types of Legitimizing Actions
We identify two broad means—external and internal—
through which new ventures can gain legitimacy in the eyes
of stakeholders (see Figure 1). Among internal means of
gaining legitimacy, we propose four types of actions: his-
torical, scientific, market, and locational. Actions associated
with historical legitimacy convey to stakeholders informa-
tion about the new ventures’ past business performance and,
by inference, their prospects for future performance. A new
60 / Journal of Marketing, July 2008
FIGURE 1
Dimensions of Legitimacy
1
We do so in part because of empirical considerations. In our
empirical analysis, we use alliances with large pharmaceutical
firms as our measure of external legitimacy. Such firms tend to
possess all four types of internal legitimacy. Therefore, a new ven-
ture’s association with such firms provides it with legitimacy on
all these types. As such, although we highlight multiple dimen-
sions of internal legitimacy, we focus here on a single dimension
of external legitimacy.
venture may convey this through, for example, a record of
product introductions, which speaks to its understanding of
the relevant technology and market. Note that this type of
legitimacy is rare, given that, by definition, few new ven-
tures have successfully introduced new products in the past.
Actions associated with scientific legitimacy convey to
stakeholders that the new ventures in question have the
technological capabilities needed to operate in their indus-
try successfully. New ventures may achieve this type of
legitimacy by, for example, recruiting eminent scientists to
serve on their boards. Actions associated with market legiti-
macy convey to stakeholders that the new ventures in ques-
tion have the market-based capabilities needed to operate in
their industry effectively. New ventures may achieve this
type of legitimacy by, for example, placing on their boards
executives who have experience in more established indus-
tries. Actions associated with locational legitimacy convey
to stakeholders that the new ventures in question derive dif-
ferential advantage as a result of their geographic location.
For example, a new software venture may do so by locating
itself in a geographic “cluster” of firms in the same special-
ized industry—for example, Silicon Valley (Porter 1998).
In contrast to the several ways new ventures may gain
internal legitimacy, we identify one major way they can
gain external legitimacy—namely, through their association
with successful and established external entities (see Rao,
Qu, and Ruekert 1999).
1
For example, a new venture may
acquire this type of legitimacy by forming an alliance with
a firm in a related but established industry, thus gaining
immediate access to the internal legitimacy that the estab-
lished firm already possesses. We now develop hypotheses
that link the dimensions of legitimacy to the rewards to
product introduction.
Hypotheses
External Legitimacy: Reflecting the Glory of the
Partner
A seemingly easy way for new ventures to gain external
legitimacy is through a formal alliance with an established
entity (Gans and Stern 2003). However, alliances have
downsides too. An alliance could result in the loss of
decision-making control and flexibility for the fledgling
venture (Das and Teng 2000). An alliance may also lead to
lower rewards to the new venture from its product introduc-
tions as a result of sharing collaborative gains (Gulati
1998). Because the new venture is likely to be smaller than
the alliance partner and thus have less bargaining power, a
larger share of the gains from product introduction may go
to the latter (Lerner and Merges 1998).
Nevertheless, an alliance with an established player can
confer benefits on new ventures in at least two ways. First,
an alliance can suggest that through its partner, the new
venture has access to the capabilities and resources needed
for successful product introduction. These would include
marketing, scientific, and financial resources, as well as
shared learning and skills (Swaminathan and Moorman
2002). Therefore, an alliance with a successful partner helps
the new venture overcome stakeholder uncertainty about its
capabilities and its inexperience in launching similar prod-
ucts. Second, the mere existence of the alliance carries its
own endorsement. That the new venture was successful in
attracting a larger, more established player in the first place
suggests that the new venture and its new product have
potential.
For these reasons, alliances can serve as a legitimizing
force that is powerful enough to overcome the negative
effects that ensue from the loss of decision-making control
and the dilution of rewards due to gain sharing. Thus, we
hypothesize the following:
The Fruits of Legitimacy / 61
H
1
: New ventures that introduce products in alliance with
established firms gain more from these product introduc-
tions than new ventures that do so without such alliances.
Internal Legitimacy, External Legitimacy, or Both?
Although we identify both external and internal means of
gaining legitimacy, we posit that firms that already have
internal legitimacy will gain fewer marginal benefits from
also acquiring external legitimacy. We do so for two rea-
sons. First, as we stated previously, alliances are not cost
free. Despite the benefits of alliances, new ventures in
emerging industries that create alliances may lose control
over their products and end up diluting future rewards due
to gain sharing (Gulati 1998; Lerner and Merges 1998;
McGarry et al. 2001). Second, new ventures that already
possess internal legitimacy may find that alliances with
established firms can lead to duplication in sources of legit-
imacy. For a new venture with internal legitimacy to form
an alliance would be akin to a manufacturer using two sub-
stitutes in a production process and paying for both when
using just one would suffice.
For the reasons we noted previously, we argue that new
ventures that possess internal legitimacy will gain more
from their products by going alone than by forming
alliances. We also note that though having both internal and
external legitimacy yields decreasing marginal benefits, the
different types of internal legitimacy actually complement
one another because they convey distinct capabilities (e.g.,
scientific legitimacy conveys technological capability, mar-
ket legitimacy conveys market knowledge). We now
develop hypotheses about how each means of gaining inter-
nal legitimacy affects the rewards to product introduction in
the presence or absence of alliances with established firms.
Historical Legitimacy: Success Speaks for Itself
Perhaps the most convincing way a new venture can gain
internal legitimacy is by developing a history of successful
product launches. This historical legitimacy confers at least
two benefits to the new venture. First, it suggests that the
new venture has prior experience with product launch, thus
increasing the likelihood of its success with future products.
Second, the large literature on “learning by doing” and
adaptive learning shows that there are clear gains to learn-
ing through a process of trial and error (Van de Ven and Pol-
ley 1992).
Thus, new ventures with historical legitimacy can gain
more from new product introduction. However, this effect
will be moderated by whether new ventures with historical
legitimacy also form alliances with established firms to
introduce new products. Firms without internal historical
legitimacy have few other means of gaining such legitimacy
than by forming alliances; for these new ventures, alliances,
though costly, may confer benefits that they would not have
otherwise. However, for new ventures that already possess
historical legitimacy, forming an alliance is likely to (1)
confer fewer marginal benefits beyond those that its experi-
ence and knowledge already provide and (2) do so at con-
siderable additional cost. Thus, new ventures that already
possess historical legitimacy are likely to be better off not
forming alliances with established firms.
H
2
: The greater a new venture’s historical legitimacy, the
larger are its gains from introducing a product on its own
than from doing so through an alliance with an established
firm.
Scientific Legitimacy: Being on the Cutting Edge
New ventures in emerging industries must convey to stake-
holders that they understand and can work with the latest
scientific ideas in the field. One way for new ventures to
achieve this scientific legitimacy is by assigning leadership
roles to scientists and academics (Stuart, Hoang, and
Hybels 1999). Not only do scientists help invent new prod-
ucts, but their presence in leadership roles in the new ven-
ture also confers at least three other benefits to the firm: (1)
technical credibility, suggesting that the new venture has the
ability within itself to develop successful new products
(Zucker, Darby, and Brewer 1998); (2) access to outside
knowledge the firm does not possess and cannot develop on
its own; and (3) some assurance that the firm has the ability
to absorb and leverage new knowledge acquired from out-
side entities (Cohen and Levinthal 1990).
As with historical legitimacy, the preceding arguments
suggest that actions associated with scientific legitimacy
can yield a substantial financial payoff for new ventures that
introduce new products. However, again, we expect that this
effect will be moderated by whether the new venture forms
an alliance with an established firm for new product intro-
duction. New ventures that do not have internal scientific
legitimacy have relatively few options to gain such legiti-
macy other than through alliances. Thus, though costly,
alliances can confer legitimacy on such new ventures that
they would not otherwise have; in turn, this would manifest
greater gains from new product introduction. In contrast, for
new ventures that already possess scientific legitimacy, also
forming an alliance would result in decreasing marginal
benefits at considerable additional cost. Thus, we hypothe-
size the following:
H
3
: The greater a new venture’s scientific legitimacy, the
larger are its gains from introducing a product on its own
than from doing so through an alliance with an established
firm.
Market Legitimacy: Knowing the Market
In addition to communicating their scientific capabilities,
new ventures in emerging industries must convey their abil-
ity to market new products. A way for new ventures to
acquire such market legitimacy is by hiring executives with
marketing and management experience in established,
related fields (Gulati and Higgins 2003). The presence in
top management positions of experienced professionals
from related fields confers three benefits on new ventures.
First, their presence suggests to important stakeholders that
the new venture understands customers well and that the
firm is market oriented in its approach. Important stake-
holders will recognize this and positively view the new ven-
tures that are market oriented. Second, because of their past
experience with commercializing new products, experi-
enced executives will provide new ventures with this impor-
tant capability, thus improving stakeholder perceptions of
the firm’s likely success with its own products. Finally, such
62 / Journal of Marketing, July 2008
executives will be able to draw on their experience with
picking market winners, again increasing the new venture’s
potential performance with its new products. Greater mar-
ket knowledge confers a greater probability of success with
new products (Li and Calantone 1998).
As with historical and scientific legitimacy, we expect
that the positive impact of actions associated with market
legitimacy will be moderated by whether new ventures that
possess it enter into alliances with established firms for new
product introduction. New ventures without market legiti-
macy will gain such legitimacy from forming alliances with
established firms; these benefits are likely to outweigh the
additional costs due to gain sharing and result in higher
returns to product introduction. Conversely, new ventures
that possess market legitimacy will gain little additional
legitimacy from forming alliances and will incur the addi-
tional costs due to gain sharing and loss of control. Conse-
quently, such firms will be better off not forming alliances
for their new product introductions. Thus, we hypothesize
the following:
H
4
: The greater a new venture’s market legitimacy, the larger
are its gains from introducing a product on its own than
from doing so through an alliance with an established
firm.
Locational Legitimacy: Where Does the Firm
Come From?
There has been considerable interest recently in the role of
“clusters” or “hot spots” in the growth of firms in emerging
industries. The research on clusters suggests that new ven-
tures in emerging industries can gain legitimacy by locating
themselves in areas with large numbers of related firms
(Pouder and St. John 1996). The presence of related firms
can help new ventures in these clusters gain such legitimacy
in three ways. First, their presence in the cluster indicates to
stakeholders that they have access through the cluster to
specialized inputs and a skilled labor pool (Porter 1998).
Second, it suggests that the new ventures have access to
new technical and market knowledge as well as resources
that can be shared (e.g., through access to university
research developed in the area). Third, it suggests that the
new ventures can potentially gain from any complementari-
ties that may result from working with or in the presence of
related firms.
There is some evidence that being in a cluster may also
convey negative information about firms located there. For
example, with time, inertia tends to set in among firms in
clusters, resulting in groupthink, nonreceptivity to external
ideas, and the suppression of innovation (Pouder and St.
John 1996). All this leads to a negative perception about
such firms and, in the process, creates an unfavorable
impression about the products these firms introduce (Gane-
san, Malter, and Rindfleisch 2005). However, because such
decline usually sets in over time and tends to afflict more
stable industries, we argue that new ventures in emerging
industries are likely to benefit rather than suffer from their
presence in clusters.
For reasons identical to those we noted for the other
forms of internal legitimacy, we expect that the positive
impact of locational legitimacy will be moderated by
whether new ventures that possess it enter into alliances
with established firms for new product introduction. Thus:
H
5
: The greater a new venture’s locational legitimacy, the
larger are its gains from introducing a product on its own
than from doing so through an alliance with an established
firm.
Method
Empirical Context
A proper test of our hypotheses requires us to investigate an
emerging industry with (1) multiple comparable product
introductions by multiple new ventures, (2) accurate details
on these product introductions, and (3) adequate variation
in the sources of internal and external legitimacy. The U.S.
biotechnology industry is an ideal industry to study for sev-
eral reasons. First, it is a truly important industry, with
applications in health care, agriculture, and industrial prod-
ucts. Biotechnology has made enormous strides in the last
two decades, with dramatic growth, considerable entry, and
equally astonishing exit rates.
Second, new products are of central importance to the
industry. In the pharmaceutical context, biotechnology rep-
resents an entirely new way of discovering drugs. As of
2002, the Food and Drug Administration (FDA) has
approved 165 products for launch, and each product has its
own unique date of introduction. Moreover, multiple firms
that fit our definition of new ventures in an emerging indus-
try have introduced these drugs. For an illustration of the
highly fragmented nature of the industry, with few clearly
dominant players in terms of product introductions to date,
see Figure 2.
Finally, as with all emerging industries, the biotechnol-
ogy industry in its early years involved much uncertainty
about its potential. This created a liability of newness for
new ventures in the industry, which in turn made it crucial
for these firms to gain legitimacy through various means in
the eyes of stakeholders. Our data show that there is suffi-
cient variation among biotech firms in the types of
approaches they adopted to gain such legitimacy (see Table
1). In summary, the industry provides a unique setting in
which to address our research questions.
Variables and Data
The stock market–based, event study method we adopt
(Brown and Warner 1985) offers an attractive approach to
estimating the gains from new products due to various legit-
imizing strategies. The method provides a quantitative mea-
sure of how the market views a particular event (e.g., a new
product launch) by computing the abnormal returns to the
firm’s stock price generated as a result of the event. We then
use this abnormal-returns measure to compute the change in
market capitalization (?MktCap) associated with the
expected rewards to product introductions by new biotech
ventures. Because stakeholders have already incorporated
information on various types of legitimacy before the event
under study (i.e., the product launch), any significant asso-
ciation between these variables and the ?MktCap of prod-
uct launch provides an appropriate test of our hypotheses.
The Fruits of Legitimacy / 63
FIGURE 2
Distribution of Product Approvals Among Biotech Firms
2
For reasons of comparability, our sample includes only biotech
drugs and not other products, such as biotech-related medical
devices.
We take a census of all biotech drugs approved by the
FDA until 2002, starting with the first approval in 1982 for
the insulin drug Novolin.
2
We focus exclusively on human
drugs for the sake of comparability and internal consistency.
Because our focus is on new ventures, we only use products
introduced by biotechnology firms and do not include
biotech drugs introduced by pharmaceutical firms. In addi-
tion, because we use stock market data to infer the rewards
associated with product introductions, we drop private
biotech firms and subsidiaries of pharmaceutical firms from
our data. Doing so yields 93 product introductions, and we
use data on these drugs in all our analyses. We now describe
each of the empirical measures we use to test our hypothe-
ses (see Table 1).
Dependent variable: rewards to product introduction.
We use abnormal stock market returns to infer the value that
stakeholders expect a new product will add to the new ven-
ture introducing it. We calculate abnormal returns over a
three-day window and use these to compute a measure of
?MktCap based on the total gains to the firm’s market capi-
tal relative to any gains to the overall market (Fama 1970)
(see Appendix A). The abnormal returns are constructed as
the difference between the return on the stock of the firm in
question and the return on the market index; this difference
controls for general movement in the market. In other
words, an event is deemed to have positive abnormal returns
if and only if the returns to it are larger than those to the
market index. Our dependent variable calculates the
absolute dollar value for rewards to product introduction
using abnormal returns. In doing so, we follow other
researchers who have also compared across many firms the
gains (or losses) that accrue to the firm from an important
event (Dowdell, Govindaraj, and Jain 1992; Hendricks and
Singhal 1997; Sorescu, Chandy, and Prabhu 2003).
3
An alternative measure of historical legitimacy would be the
entrepreneurial history of the top managers of the firm. However,
only 9.79% of top managers or board members in our sample had
any history of entrepreneurship. We reestimate our models using
this measure but find that its coefficients are not significantly dif-
ferent from zero in any of the models.
Independent variables: external and internal legitimacy.
To construct measures of our independent variables, we use
publicly available information, including annual reports,
company Web sites, and trade journals. We explain these
measures and their sources in detail subsequently.
To measure actions associated with external legitimacy,
we use data on alliances between biotech firms in our sam-
ple and established pharmaceutical firms. We collect these
data from trade journals such as Pharma Business Week and
press reports. We code a dummy variable on the basis of
whether the new venture introduced the product in alliance
with an established pharmaceutical firm. Established phar-
maceutical firms are those that had a sizable presence in
the pharmaceutical business before 1976 (a year widely
regarded as the birth of the first biotechnology firm Genen-
tech). In practice, these firms are mostly members of “big
pharma,” the largest pharmaceutical firms worldwide.
Because our hypotheses on internal legitimacy (H
2
–H
5
)
are interactions between the lack of an alliance and the dif-
ferent sources of internal legitimacy, we reverse-code the
alliance variable for easier interpretation. We code the
variable as 1 if the biotech firm has no alliance with a phar-
maceutical firm (i.e., the new venture “goes alone”) and 0 if
the firm has such an alliance.
We measure internal legitimacy in terms of four compo-
nents: historical, scientific, market, and locational legiti-
macy. We measure actions associated with historical legiti-
macy by identifying the number of drug approvals the
biotech firm obtained before the focal approval.
3
Drug
approval in the pharmaceutical industry is a long and
expensive process. The presence of one or more approved
drugs in a biotech new venture’s portfolio provides immedi-
64 / Journal of Marketing, July 2008
Conceptual Variable Notation Measured Variable Data Sources
Rewards to product
introduction
?MktCap
it
Net present value in 1987 dollars
($ millions)
•Center for Research in Security
Prices
•Dow Jones/Reuters
External legitimacy (1 – GoAlone
ijt
) Alliance with established
pharmaceutical firms
•Pharma Business Week
•Med Ad News
•www.bio.org
Historical legitimacy Hist
it
The number of previous product
approvals
•Pharma Business Week
•Med Ad News
•www.bio.org
Scientific legitimacy Sci
it
Fraction of board consisting of
academicians
•LexisNexis
•SEC filings
Market legitimacy Mkt
it
Fraction of board consisting of
ex-pharmaceutical executives
•LexisNexis
•SEC filings
Locational legitimacy Loc
it
The number of biotech firms in the
city area
•CorpTech database
Technology know-how KnowHow
it
The stock of patents for focal firm on
approval date
•Delphion
Firm size Size
it
Log of number of employees •COMPUSTAT
Age Age
it
Age of the firm •Annual reports
•Company Web sites
Consumer benefits Priority
jt
Priority status granted by FDA •Pharmaprojects
•www.fda.gov
Niche product Orphan
jt
Orphan status granted by FDA •Pharmaprojects
•www.fda.gov
Market potential MktPot
jt
Estimated market potential of the
therapeutic category
•Pharmaprojects
Competitive intensity Comp
jt
Number of existing players in the
therapeutic category
•Pharmaprojects
Alliance dates Date of formation of product alliance •Recap database
TABLE 1
Variables and Data Sources
ate testimony to stakeholders of its ability to use resources
effectively and of its experience in developing and market-
ing drugs. This testimony is particularly strong given that
there were just 165 approved biotech drugs on the market as
of 2002; the number of biotech firms at that time was more
than 1500. Having an approved drug in its portfolio, there-
fore, places a biotech new venture in an exclusive club. As
Figure 2 shows, a few firms have one or more prior product
introductions, though most firms do not. Only one firm—
Genentech—has as many as 10 products approved, 4 more
than its nearest rival, Amgen.
4
4
This may suggest that Genentech is an outlier in our model.
However, an outlier analysis indicates that this is not the case. In
addition, when we include a dummy variable for Genentech to
capture its fixed effects in our subsequent analyses, we find that its
We measure actions associated with scientific legiti-
macy by computing the ratio of academic scientists on the
biotech firm’s board to the total size of the board. We count
as an academic scientist anyone who holds an appointment
in a university as a researcher or a faculty member. A high
ratio of academic scientists on the board conveys to stake-
holders that the firm has the technical capabilities needed
for successful product introduction (Dacin 1997). Stake-
holders are likely to respond to the prominence of scientists
on the firm’s board. Thus, two scientists on a four-member
board will convey more scientific legitimacy than two sci-
entists on a ten-member board. We obtain data on the com-
position of corporate boards and on board members’ affilia-
coefficient is not significantly different from zero, and the other
results remain essentially unchanged.
The Fruits of Legitimacy / 65
5
Ideally, we would have a time-varying measure of this variable.
Because of data availability constraints, however, we construct this
variable for 1994 and use this measure for each period thereafter
with the assumption that relative cluster density remains the same
for preceding and succeeding years.
tions with research institutions from Securities and
Exchange Commission (SEC) filings.
We measure actions associated with market legitimacy
by computing the ratio of former or current executives of
established pharmaceutical firms on the biotech firm’s
board to the total size of the board. A high ratio of such
executives on the board conveys to stakeholders that the
firm has the market capabilities needed for successful prod-
uct introduction (Aldrich and Baker 2001). We obtain infor-
mation on board members’ biographies and prior affiliations
from SEC filings.
We measure actions associated with locational legiti-
macy by calculating the number of biotech firms in the met-
ropolitan statistical area in which the focal firm is located.
The higher the number of such firms in the city area, the
closer it resembles a cluster. We also construct two alterna-
tive measures of locational legitimacy. First, we compute a
dummy variable and assign new ventures in well-known
biotech “hot spots” (i.e., California, Massachusetts, Mary-
land, and New York/New Jersey) a value of 1 and new ven-
tures in other areas a value of 0. Second, we develop a state-
wise measure of “clustering” by computing the ratio of
biotech firms in the state of the focal firm divided by the
total number of firms in the United States.
5
As we report
subsequently, our results are robust to the use of all these
measures of locational legitimacy.
Control variables. Although our theory predicts a sig-
nificant relationship between the legitimacy variables and
the rewards to product introduction, several other firm- and
product-specific factors are also likely to influence the
?MktCap of new products. We obtain measures of these
control variables as follows.
For firm-specific controls, we consider firm size, firm
age, and technological know-how. First, larger firms are
likely to have greater resources that they can leverage to
introduce products more successfully. Although the evi-
dence on this is mixed (see Sorescu, Chandy, and Prabhu
2003), we control for size to capture the heterogeneity
among firms in our sample. We construct a measure of firm
size using the natural log of the number of employees in the
firm in the year of approval. Second, we control for the age
of the firm to tease out the effect of legitimacy from the
pure time effect; that is, as firms get older, they might
become more established, and this might influence the
returns associated with a product launch (Hannan and Free-
man 1984). Accordingly, we control for firm age, measured
as the number of years since the focal firm’s incorporation.
Third, because some of our measures of legitimacy might
be considered merely proxies for the technological know-
how of new ventures, we explicitly account for technologi-
cal know-how by including as a control variable the stock
of patents for the new venture on the date of product intro-
6
We also performed the analysis using an alternative measure of
technological know-how—the number of patents weighted by cita-
tion counts. Our hypothesis tests yield the same results.
duction. This measure has been frequently used in prior
studies to measure technological know-how (see Prabhu,
Chandy, and Ellis 2005).
6
To estimate the effects of legitimacy on the rewards to
product innovation, we need to account for the inherent
quality of each product and the market context in which it is
introduced. We do so by controlling for the following fac-
tors: consumer benefits relative to existing products, a priori
market potential, niche products, and competitive intensity.
First, we code whether the FDA deemed the drug to provide
substantially higher therapeutic benefits relative to existing
drugs and therefore granted it priority review status
(Sorescu, Chandy, and Prabhu 2003). Second, we obtain an
estimate of market potential (in billions of dollars) for each
therapeutic category from the Pharmaprojects database.
This variable accounts for the notion that, for example, an
AIDS drug faces inherently different demand conditions
than a dermatology drug. Third, as an additional measure of
market potential, we code whether the focal drug is an
orphan (i.e., a drug for which the target population is lim-
ited or for which the disease it treats occurs only rarely).
Granting of orphan status by the FDA confers exclusive
access to the market for the drug. Fourth, we account for
competitive intensity by counting the number of competi-
tors that are already in the therapeutic category.
Model
Our unit of analysis is the individual product introduction
(i.e., FDA approval for market launch). We use the ?Mkt-
Cap associated with each product approval as the dependent
variable. Each ?MktCap is associated with a firm and a
time period. However, because there are few multiproduct
biotech firms (see Figure 2), applying a panel specification
to the data is difficult. Instead, we attempt to parcel out
alternative sources of variance in the data by including a
large number of control variables in our model
specification.
We specify the following model for testing our hypothe-
ses using the ?MktCap generated on the stock market as the
dependent variable and the legitimacy measures as the inde-
pendent variables, along with the firm- and product-specific
control variables described previously:
(1) ?MktCap
it
= ?
0
+ ?
g
GoAlone
ijt
+ ?
gh
Hist
it
× GoAlone
ijt
+ ?
gs
Sci
it
× GoAlone
ijt
+ ?
gm
Mkt
it
× GoAlone
ijt
+ ?
gl
Loc
it
× GoAlone
ijt
+ ?
gk
KnowHow
it
× GoAlone
ijt
+ ?
ah
Hist
it
× (1 – GoAlone
ijt
)
+ ?
as
Sci
it
× (1 – GoAlone
ijt
)
+ ?
am
Mkt
it
× (1 – GoAlone
ijt
)
+ ?
al
Loc
it
× (1 – GoAlone
ijt
)
66 / Journal of Marketing, July 2008
+ ?
ak
KnowHow
it
× (1 – GoAlone
ijt
)
+ ?
1g
Age
it
× GoAlone
ijt
+ ?
2g
Size
it
× GoAlone
ijt
+ ?
3g
MktPot
jt
× GoAlone
ijt
+ ?
4g
Comp
jt
× GoAlone
ijt
+ ?
5g
Orphan
jt
× GoAlone
ijt
+ ?
6g
Priority
jt
× GoAlone
ijt
+ ?
1a
Age
it
× (1 – GoAlone
ijt
)
+ ?
2a
Size
it
× (1 – GoAlone
ijt
)
+ ?
3a
MktPot
jt
× (1 – GoAlone
ijt
)
+ ?
4a
Comp
jt
× (1 – GoAlone
ijt
)
+ ?
5a
Orphan
jt
× (1 – GoAlone
ijt
)
+ ?
6a
Priority
jt
× (1 – GoAlone
ijt
) + ?
ijt
,
where i, j, and t are subscripts for firm, product, and time,
respectively, and the other variables are defined in Table 1.
We expect ?
g
to be negative and ?
gh
, ?
gs
, ?
gm
, and ?
gl
to be
positive per hypotheses H
1
, H
2
, H
3
, H
4
, and H
5
, and (1 –
GoAlone
ijt
) is 1 when a firm has an alliance (i.e., it does not
go alone). The model is interactive in the sense that each of
the independent variables can affect the dependent variable
differently, depending on whether the product in question
involves an alliance (i.e., the regression coefficients are dif-
ferent under each regime).
The specification in Equation 1 can be estimated using
standard ordinary least squares (OLS) regression. However,
doing so could be problematic because a firm’s decision to
go alone or form an alliance is a self-selected variable. For
example, a firm might decide to go alone because the prod-
uct it is seeking approval for has a quality advantage; there-
fore, the firm might expect a large revenue stream to accrue
from the product and thus might not be willing to share the
revenue with an alliance partner. Nevertheless, this unob-
served quality advantage (possibly observed by the market
but not by the econometrician) might also affect the ?Mkt-
Cap returns to the firm from product introduction. There-
fore, we have a classic case of endogeneity in which an
independent variable is correlated with the error term, and
thus estimation using OLS would violate the fundamental
OLS assumption of independent errors (Greene 2003).
7
Econometrically, the way to correct for endogenous
regressors is to use an instrumental variables approach. In
our case, the endogenous variable—the choice to go alone
7
It could be argued that the other legitimacy variables in Equa-
tion 1 are also endogenous. We focus on the alliance variable only
and consider other variables exogenous because these are not
product-specific decisions, whereas an alliance is. Thus, although
the alliance decision might be driven in part by the unobserved
quality of the product and thus might be correlated with it, this is
not likely to be the case for the other measures of legitimacy,
because these are more long-term, firm-related (rather than
product-related) measures. We address this issue in detail
subsequently.
or to form an alliance—is a dummy variable. Therefore, we
need to use a slightly different approach to correct for
enodogeneity than the standard instrumental variables
approach.
In our case, the decision of whether to form an alliance
is similar to what applied economists call assignment to a
“treatment” (for a review, see Angrist 2001). The decision
to form an alliance or go alone in launching a product is
undertaken with some expectations regarding the outcome
of the decision. A common class of econometric models
termed “treatment effect” models has been developed to
overcome the bias introduced by self-selected (endogenous)
dummy variables. In these models, the endogenous variable
is a dummy defined by whether a person was assigned to
a “treatment” (Angrist 2001; Greene 2003; Maddala 1983).
In our case, however, we cannot directly apply a treatment
model because the treatment (i.e., the decision to go alone)
also appears in the interaction terms in Equation 1. To over-
come this problem, we employ the following estimation
strategy: We run two separate regressions—one for firms
that introduce products in an alliance and one for firms that
choose to go alone. For firms that go alone, Equation 1
reduces to the following:
(2) ?MktCap
it
= (?
0
+ ?
g
) + ?
gh
Hist
it
+ ?
gs
Sci
it
+ ?
gm
Mkt
it
+ ?
gl
Loc
it
+ ?
gk
KnowHow
it
+ ?
1g
Age
it
+ ?
2g
Size
it
+ ?
3g
MktPot
jt
+ ?
4g
Comp
jt
+ ?
5g
Orphan
jt
+ ?
6g
Priority
jt
+ ?
ijt
.
For firms that form an alliance, Equation 1 reduces to the
following:
(3) ?MktCap
it
= ?
0
+ ?
ah
Hist
it
+ ?
as
Sci
it
+ ?
am
Mkt
it
+ ?
al
Loc
it
+ ?
ak
KnowHow
it
+ ?
1a
Age
it
+ ?
2a
Size
it
+ ?
3a
MktPot
jt
+ ?
4a
Comp
jt
+ ?
5a
Orphan
jt
+ ?
6a
Priority
jt
+ ?
ijt
.
Now, the estimation of parameters of interest is made easier.
The difference in the estimated intercepts of Equations 2
and 3 gives an estimate of ?
g
, the difference in the estimated
coefficient of Hist in Equations 2 and 3 gives an estimate of
?
gh
– ?
ah
, and so on.
However, note that an OLS-only estimation of Equa-
tions 2 and 3 would not solve the problem of self-selection
(Shaver 1998).
8
To obtain unbiased estimates of Equations
2 and 3, we run each regression separately and correct for
the self-selection problem (Heckman 1979; Shaver 1998).
We assume that all the variables that drive ?MktCap also
drive firms’ decision to go alone or form an alliance. For
example, the firms with greater internal legitimacy might be
less likely to form an alliance because they will have greater
confidence in their ability to go alone (Eisenhardt and
Schoonhoven 1996; Lerner and Merges 1998). The under-
8
Despite this, we also estimate these equations using OLS and
find that, in general, the OLS results are similar to the results from
the models we report and discuss subsequently.
The Fruits of Legitimacy / 67
lying equations that drive the selection (i.e., to go alone or
form an alliance) in Equations 2 and 3 are as follows:
(4a) GoAlone
ijt
= µ
0
+ µ
1
Hist
it
+ µ
2
Sci
it
+ µ
3
Mkt
it
+ µ
4
Loc
it
+ µ
5
KnowHow
it
+ µ
5
Age
it
+ µ
7
Size
it
+ µ
8
MktPot
jt
+ µ
9
Comp
jt
+ µ
10
Orphan
jt
+ µ
11
Priority
jt
+ ?
ijt
, and
(4b) Alliance
ijt
= µ
0
?+ µ
1
?Hist
it
+ µ
2
?Sci
it
+ µ
3
?Mkt
it
+ µ
4
?Loc
it
+ µ
5
?KnowHow
it
+ µ
5
?Age
it
+ µ
7
?Size
it
+ µ
8
?MktPot
jt
+ µ
9
?Comp
jt
+ µ
10
?Orphan
jt
+ µ
11
?Priority
jt
+ ?
ijt
?.
To test our hypotheses formally, we use maximum like-
lihood estimation to jointly estimate Equations 2 and 4a and
Equations 3 and 4b, respectively, while accounting for the
dichotomous nature of the dependent variables in Equations
4a and 4b (see Appendix B). Note that the dependent
variable in Equation 4a is coded as 1 when a firm goes
alone and as 0 when it has an alliance. Conversely, the
dependent variable in Equation 4b is coded as 0 when a
firm goes alone and as 1 when it has an alliance. To clarify
this point, consider the joint estimation procedure for Equa-
tions 2 and 4a. When these two equations are estimated
jointly, the ?MktCap values for all firms that go alone are
treated as observed, whereas those for firms that form
alliances are treated as censored. Likewise, when Equations
3 and 4b are estimated jointly, the ?MktCap values for all
firms that form alliances are treated as observed, whereas
those for firms that go alone are treated as censored.
Results
Table 2 presents descriptive statistics for our data. Note that
the average ?MktCap across all firms is $62 million (in
1987 dollars); given that the average market capitalization
across all firms before the FDA approval is $1.47 billion,
this represents an average gain of approximately 4.2%.
Table 3 exhibits the parameter estimates for the models.
Detailed analyses of the influence of outliers, nonnormality,
and multicollinearity indicate that these do not affect the
results of the estimation. As we mentioned previously, a
simple OLS specification of our model provides results
consistent with our key hypotheses. However, Table 3
shows that the estimated coefficient (?) for the selection
correction term (Mills ratio) is significantly different from
zero in both the go-alone and the alliance regimes (–837.77,
p < .01, and 189.27, p < .01, respectively). These results
imply that there is indeed an underlying selection process in
firms’ decisions to go alone or to form an alliance, and thus
a model that accounts for the underlying selection process
provides more consistent and unbiased results than OLS.
The difference between the relevant estimates of the two
models provides parameters that show how the various
types of actions associated with legitimacy differentially
9
Specifically, if we denote the estimates for the go-alone model
as 1 and those for the alliance model as 2, then ?
new
= ?
n1
– ?
n2
,
?
new
= ??
2
n1
+ ?
2
n2
, and DF
new
= (?
2
n1
+ ?
2
n2
)
2
/(?
4
n1
/DF
n1
+
?
4
n2
/DF
n2
), where the first equation represents the difference in
coefficients obtained in the individual regressions, the second
equation represents the standard error of the coefficient obtained
after differencing, and the third equation represents the degrees of
freedom used to calculate the t-statistics of the differenced coeffi-
cients (Moore and McCabe 1998; Satterthwaite 1946).
10
To test whether the underlying model of Equation 1 with
interaction effects is indeed the “true model,” we run a simple
main-effects version of Equation 1 with no interaction terms. We
then run a Chow test using the parameters from this model and the
parameters from Equations 2 and 3 in Table 3. The Chow test
reveals that the coefficients of the two models in Table 3 are
indeed different (F = 20.86, p < .01), confirming that interaction
effects exist.
affect the ?MktCap in the GoAlone versus Alliance cases,
respectively.
9
Table 4 lists these values.
10
Effects of External Legitimacy: Do Alliances
Matter?
H
1
predicts that new ventures that introduce products in
alliance with established firms gain more from these prod-
uct introductions than new ventures that do so without such
alliances. As Table 4 shows, the coefficient of the GoAlone
dummy is negative and significant (?
g
= –300.95, p < .1), in
support of H
1
. Thus, we find support for the positive effects
of external legitimacy on rewards to product introduction by
new ventures in an emerging industry.
Effects of Internal Legitimacy
Historical legitimacy. H
2
predicts that, all else being
equal, the greater a new venture’s historical legitimacy, the
larger are its gains from introducing a product without an
alliance than through an alliance with an established firm.
The positive, significant coefficient of ?
gh
– ?
ah
(287.61,
p < .01; see Table 4) strongly supports this hypothesis. The
coefficients in Table 3 provide additional details regarding
this result. As Model 1 in Table 3 shows, for firms that go
alone, greater historical legitimacy results in greater gains
from new product introductions (?
gh
= 295.37, p < .01). In
contrast, Model 2 shows that for firms that form alliances,
historical legitimacy has a nonsignificant effect on gains
from product introductions. These results suggest that when
a firm forms an alliance, the marginal benefits of historical
legitimacy are absent.
Scientific legitimacy. H
3
predicts that, all else being
equal, the greater a new venture’s scientific legitimacy, the
larger are its gains from introducing a product without an
alliance than through an alliance with an established firm.
The positive, significant coefficient of ?
gs
– ?
as
(1070.52,
p < .1; see Table 4) supports this hypothesis. Again, the
coefficients in Table 3 provide details regarding this result.
As Model 1 in Table 3 shows, for firms that go alone,
greater scientific legitimacy results in greater gains from
new product introductions (?
gs
= 1369.51, p < .1). In con-
trast, Model 2 shows that for firms that form alliances, sci-
68 / Journal of Marketing, July 2008
T
A
B
L
E

2
D
e
s
c
r
i
p
t
i
v
e

S
t
a
t
i
s
t
i
c
s

a
n
d

C
o
r
r
e
l
a
t
i
o
n

M
a
t
r
i
x
M
S
D
?
M
k
t
C
a
p
A
l
l
i
a
n
c
e
H
i
s
t
S
c
i
M
k
t
L
o
c
K
n
o
w
H
o
w
A
g
e
M
k
t
P
o
t
C
o
m
p
S
i
z
e
P
r
i
o
r
i
t
y
O
r
p
h
a
n
?
M
k
t
C
a
p
6
2
.
1
3
3
9
8
.
2
4
1
.
0
0
A
l
l
i
a
n
c
e
.
4
2
.
5
0
–
.
0
5
1
.
0
0
H
i
s
t
1
.
0
8
1
.
6
8
.
3
9
*
.
0
5
1
.
0
0
S
c
i
.
2
0
.
1
5
.
1
8
*
.
0
9
.
0
7
1
.
0
0
M
k
t
.
2
1
.
1
7
.
1
1
.
0
4
–
1
.
0
7
–
.
0
3
1
.
0
0
L
o
c
1
2
.
1
7
1
3
.
8
.
1
7
.
0
9
.
0
6
.
1
7
.
0
4
1
.
0
0
K
n
o
w
H
o
w
1
3
9
.
0
4
1
9
5
.
9
.
2
9
*
–
.
1
2
.
4
2
*
.
0
4
–
.
0
3
–
.
0
8
1
.
0
0
A
g
e
1
5
.
3
9
7
.
9
4
.
0
7
–
.
1
2
.
2
3
*
–
.
0
6
–
.
1
1
–
.
0
8
.
7
3
*
1
.
0
0
M
k
t
P
o
t
2
.
5
4
2
.
2
0
–
.
0
2
.
1
2
.
0
8
–
.
0
9
–
.
3
2
*
.
1
8
.
0
3
.
0
9
1
.
0
0
C
o
m
p
2
3
.
0
6
1
4
.
2
8
–
.
0
0
4
–
.
0
6
.
1
1
.
0
3
.
2
1
*
.
0
3
.
3
4
*
.
1
6
–
.
0
4
1
.
0
0
S
i
z
e
6
.
3
1
1
.
5
1
.
1
8
*
–
.
0
2
.
5
9
*
.
1
0
–
.
2
8
*
.
0
6
.
6
5
*
.
4
4
*
.
1
7
.
2
9
*
1
.
0
0
P
r
i
o
r
i
t
y
.
5
4
8
.
5
0
0
–
.
1
0
1
–
.
0
1
7
–
.
1
5
6
–
.
0
9
1
–
.
0
9
9
–
.
0
2
3
–
.
1
0
6
–
.
0
3
2
–
.
0
5
1
–
.
0
4
5
–
.
0
7
3
1
.
0
0
O
r
p
h
a
n
.
3
3
3
.
4
7
4
–
.
0
1
3
–
.
0
0
0
.
0
7
1
.
2
5
1
*
.
0
0
8
.
2
6
9
*
–
.
2
0
0
–
.
1
9
1
–
.
0
3
8
–
.
1
4
3
–
.
0
1
8
–
.
1
3
7
1
.
0
0
*
p
<

.
0
5
.
The Fruits of Legitimacy / 69
TABLE 3
Results of Selection Models
Variable
Model 1 (GoAlone
Firms Only)
Selection
(for Model 1)
Model 2 (Alliance
Firms Only)
Selection
(for Model 2)
Intercept –473.32**
(211.15)
.99
(1.41)
–172.37
(127.23)
–.315
(.915)
Hist 295.37***
(78.44)
.055
(.121)
7.76
(17.73)
.087
(.115)
Sci 1369.51*
(753.87)
–1.28
(.88)
299.33**
(160.14)
1.04
(.878)
Mkt 1605.65***
(647.40)
–1.95*
(.98)
88.96
(156.82)
.221
(.872)
Loc –8.87
(8.09)
.004
(.009)
3.06*
(1.80)
.018*
(.011)
KnowHow –.575
(1.00)
.0029*
(.0017)
–.100
(.222)
–.0006
(.001)
Priority 180.01
(212.23)
.007
(.257)
–45.82
(59.24)
–.321
(.297)
Orphan 2.84
(238.63)
–.089
(.276)
26.21
(66.91)
–.069
(.368)
Size –114.86
(95.00)
–.063
(.178)
164.46
(199.59)
–.046
(.150)
MktPot 129.42***
(49.33)
–.147**
(.060)
23.59**
(11.70)
.078
(.071)
Age –19.75
(19.59)
.008
(.031)
–3.06
(5.63)
–.010
(.028)
Comp 5.01
(7.31)
–.0006
(.009)
–1.59
(1.89)
–.0016
(.0099)
N: total 93 93
N: uncensored 54 39
Correction: self-selection (?) –837.77*** 0189.27***
Log-likelihood –466.71*** –287.88***
*p < .1.
**p < .05.
***p < .01.
Notes: Standard errors are in parentheses.
TABLE 4
Estimated Coefficient Differences
Expected Sign Implication
GoAlone (?
g
) –300.95*
(238.74)
Negative H
1
supported
?
gh
– ?
ah
(historical legitimacy) 287.61**
(80.41)
Positive H
2
supported
?
gs
– ?
as
(scientific legitimacy) 1070.52*
(770.69)
Positive H
3
supported
?
gm
– ?
am
(market legitimacy) 1516.69**
(666.12)
Positive H
4
supported
?
gl
– ?
al
(locational legitimacy) –11.93*
(8.28)
Positive H
5
not supported
*p < .1.
**p < .01.
Notes: Standard errors are in parentheses. Calculation of p-values is as per n. 9.
entific legitimacy has a positive but significantly lower
effect on gains from product introductions (?
as
= 299.33,
p < .05).
Market legitimacy. H
4
predicts that, all else being equal,
the greater a new venture’s market legitimacy, the larger are
its gains from introducing a product without an alliance
than through an alliance with an established firm. As Table
4 shows, ?
gm
– ?
am
is positive and significant (1516.69, p <
.01), in support of H
4
.
Locational legitimacy. Finally, H
5
predicts that, all else
being equal, the greater a new venture’s locational legiti-
macy, the larger are its gains from introducing a product
without an alliance than through an alliance with an estab-
lished firm. Contrary to H
5
, the coefficient of ?
gl
– ?
al
is
negative and significant (–11.93, p < .1). This unexpected
result suggests that the much-researched “cluster effect”
does not yield higher financial returns to product introduc-
tions in the biotechnology industry. Nevertheless, though
70 / Journal of Marketing, July 2008
negative, the overall effect size is small, suggesting that
being in a cluster does not have too negative an impact on
?MktCap.
Internal Versus External Legitimacy
Figure 3 describes the impact of forming an alliance com-
pared with going alone for firms with different levels of
internal legitimacy. We plot the estimated difference in
?MktCap (calculated using the coefficients in Table 3)
between firms with high internal legitimacy that form
alliances and those with high internal legitimacy that go
alone. High-legitimacy firms are those with each internal
legitimacy variable set to one standard deviation above the
mean value in the sample. Similarly, we calculate the esti-
mated difference in ?MktCap (again using the coefficients
in Table 3) between firms with low internal legitimacy that
form alliances and those with low legitimacy that go alone.
Low-legitimacy firms are those with each internal legiti-
macy variable set to one standard deviation below the mean
value in the sample. We find that a firm with low internal
legitimacy that goes alone gains $367 million less in ?Mkt-
Cap than one that introduces the product in alliance with an
established firm. In contrast, a firm with high internal legit-
imacy that goes alone gains $497 million more in ?MktCap
than one that introduces the product in alliance with an
established firm (see Figure 3).
Additional Analyses
Do gains from alliances occur before product introduc-
tion? A key result is that new ventures with high internal
legitimacy gain more from going alone than from forming
alliances. It might be argued that an alternative explanation
exists for this result. Specifically, firms with high levels of
internal legitimacy may be able to cut better deals when
forming a product-specific alliance before the product is
approved. For example, a firm with reputed scientists on its
board might enter alliance negotiations with greater bar-
gaining power, resulting in higher abnormal gains at the
time of alliance formation. Consequently, the ?MktCaps on
product approval for firms with high internal legitimacy that
form alliances might be low simply because the market has
already incorporated the potential gains from product intro-
duction at the time the alliance was announced.
To examine this possibility, we obtain the announce-
ment dates of all product-specific alliances of firms in our
data set. We use these dates to calculate the ?MktCaps due
to alliance announcements. We then regress these alliance-
specific ?MktCaps against the measures of internal legiti-
macy. We do not find a significant, positive association
between the legitimacy variables and the alliance-specific
?MktCaps. Furthermore, we do not find a significant corre-
lation between the ?MktCaps at the time of alliance forma-
tion and the ?MktCaps at product approval (pairwise
correlation = –.0002, p = .3). Thus, we reject the alternative
explanation that our results are driven by not taking into
account the increase in firm value at the time of alliance for-
mation. Finally, although we find that internal legitimacy
has no effect on alliance ?MktCaps, we find that the spe-
cific details of alliances (e.g., royalty terms, property
rights–sharing agreements) do have a significant effect on
them.
Does the exclusion of failed new product introductions
affect our results? Given our emphasis on the returns to new
product introductions, the results we report focus only on
FIGURE 3
Impact of Internal Versus External Legitimacy
The Fruits of Legitimacy / 71
products that are actually introduced into the marketplace.
Nevertheless, does legitimacy help or hurt the odds of prod-
ucts under development making it to launch in the first
place? To address this question, we obtain additional data
on all products—both successes and failures—that entered
the FDA drug pipeline during our study period (1982–
2002). These data cover products that were approved,
discontinued, or remained in the product development
pipeline. In all, we have 530 observations with 93
approvals, 187 rejections, and 250 with right truncation. As
such, we analyze these data using a competing risk Cox
proportional hazard model (see Lunn and McNeil 1995).
This analysis suggests that legitimacy is actually associated
with lowered odds of failure in the FDA approval process,
even after we account for investments in research and devel-
opment (R&D) and priority status in clinical trials. Specifi-
cally, a firm with historical legitimacy obtained through a
previous product introduction improves its chances of being
approved (relative to being discontinued) by approximately
26% (e
.231
? 1.26). A firm that has an external alliance with
an established firm improves its chances of being approved
(relative to being discontinued) by approximately 63%
(e
.492
? 1.63). Locational legitimacy also pays off in terms
of increased probability of approval; a one standard devia-
tion improvement in locational legitimacy leads to an
increased hazard of being approved (relative to being dis-
continued) of approximately 47%. In contrast, a high mar-
ket potential decreases the chances of being approved, per-
haps because of greater regulatory scrutiny for more
important drugs. Investments in R&D also pay off, by yield-
ing better odds of approval (or correspondingly lower odds
of failure). Similarly, receiving priority review status from
the FDA dramatically improves a drug’s chance of
approval. The results of this analysis suggest that legitimacy
is actually associated with lowered odds of failure in the
FDA approval process, even after we account for invest-
ments in R&D. In other words, legitimacy actually helps
products under development make it to launch in the first
place. These results further highlight the importance of
legitimacy in driving the success of innovations by new
ventures.
Are our results robust to alternative estimation meth-
ods? We estimate the results in Table 3 using maximum
likelihood. To check for the robustness of our results to the
estimation method, we also estimate the model using (1) an
OLS specification, (2) a Cox proportional hazard model,
and (3) a Heckman two-stage procedure. The results from
these analyses are consistent with those from the maximum
likelihood estimation procedure we report here. The consis-
tency of our results across these methods reinforces the
robustness of the results.
Do the legitimacy variables act as proxies for quality?
If the legitimacy variables we examine simply act as prox-
ies for product quality, they would not play the suggested
conceptual role in explaining returns to product introduc-
tion. Moreover, they would cause concerns about additional
endogeneity in our model and raise questions about our
results regarding the effects of legitimacy on the rewards to
product introduction because it would be unclear whether
11
None of the quality measures have a significant effect on sci-
entific legitimacy, and none of the quality measures except for the
a priori market potential measure have a significant effect on mar-
ket legitimacy.
the results reflect the true effects of legitimacy or simply the
indirect effects of product quality.
We address these concerns both conceptually and
empirically. Conceptually, a firm with a high-quality prod-
uct might be able to attract high-quality scientific and mar-
ket talent, thus increasing its scientific and market legiti-
macy. In contrast to scientific and market legitimacy,
however, locational and historical legitimacy are both
largely tied to the origins of each firm. In general, they are
established long before the introduction of a particular
product and are not easily changed in response to a high-
quality product. As such, locational and historical legiti-
macy do not serve as proxies for the quality of a particular
product.
To address the empirical concern that scientific and
market legitimacy might simply be proxies for product
quality, we run OLS regressions with scientific and market
legitimacy measures as dependent variables and various
product-, market-, and firm-specific measures of quality as
independent variables. Specifically, we include the follow-
ing independent variables: a priori market potential, relative
consumer benefits, whether the product was a niche prod-
uct, competitive intensity in the market, and a quality-
adjusted metric of each firm’s know-how that weights the
patent stock of each firm by the citations received by each
patent. The results from this analysis indicate that most of
these measures have no significant effect on either scientific
or market legitimacy.
11
These results provide some reassur-
ance that our measures of internal legitimacy do not simply
serve as proxies for quality.
Which sources of internal legitimacy matter more? The
estimates of the internal legitimacy variables (in Table 3)
provide dollar measures of the impact of the improvements
in various legitimizing strategies. For example, a one-
standard-deviation increase in historical, scientific, and
market legitimacy decreases the negative impact of going
alone by $495 million, $205 million, and $272 million,
respectively. We also estimate standardized coefficients
(from Table 3) for the internal legitimacy variables to assess
the relative impact of these variables on ?MktCap when
firms go alone. We find that historical legitimacy and mar-
ket legitimacy have the greatest impact on payoffs (z = 3.77,
p < .01, and z = 2.48, p < .01, respectively), followed by
scientific (z = 1.82, p < .1) legitimacy. As we reported pre-
viously, the impact of geographical legitimacy is negative
(z = 1.10, p < .3). These results have many implications, as
we discuss next.
Discussion and Implications
We show that new ventures can overcome the liability of
newness by adopting strategies that give them legitimacy in
the eyes of stakeholders. New ventures can gain legitimacy
72 / Journal of Marketing, July 2008
through the associations they form with other, more estab-
lished and reputed entities. We find that this legitimacy pays
off by directly raising the rewards to such firms from their
product introductions.
All else being equal, we find that new ventures that
acquire external legitimacy by forming alliances with estab-
lished firms in related industries gain more from their new
products than new ventures that do not form such alliances.
Among new ventures that do not form alliances, those that
acquire internal legitimacy by creating a history of success-
ful product introduction or by hiring reputed scientists and
managers gain more from their new products than those that
do not. Finally, although new ventures can gain from either
external or internal legitimacy, we find that pursuit of exter-
nal legitimacy by firms that already have internal legitimacy
leads to lower rewards to innovation.
Implications for Research
Our findings have three major implications for research.
First, we show that legitimacy explains the differential gains
to new ventures from new product launch. Prior research on
the issue of why some firms gain more from their new prod-
ucts than others suggests that this is due to variables such as
firm size, know-how, resources, or the attributes of the
product launched. Although these explanations are relevant
to mature industries, they are less applicable to emerging
industries in which firms are roughly the same size and new
products are all more or less breakthrough in nature. We
show that product gains vary greatly between firms that
adopt legitimizing strategies and those than do not. The lit-
erature has traditionally emphasized how a desire for legiti-
macy pushes firms toward conformity with each other. Our
results stress the need for more research on how managers
use legitimizing strategies to differentiate their firms from
others.
Second, this article is one of the first in marketing to
incorporate the role of legitimizing forces in determining
the outcomes of marketing actions (see Grewal and Dhar-
wadkar 2002). The effects of legitimacy have been studied
in other contexts, linking it to firm survival rates (Ruef and
Scott 1998), new venture growth (Zimmerman and Zeitz
2002), and initial public offering success (Higgins and
Gulati 2003). Our results are consistent with these previous
studies. However, we go beyond these studies and outline a
typology of legitimizing strategies in the context of product
introduction. The role of legitimizing strategies in other
marketing contexts is a topic that merits additional research.
Third, in contrast to much of the literature on the bene-
fits of locating new ventures in industrial clusters, we find
that firms seeking this form of legitimacy actually lose out
compared with those that do not. This finding suggests that
location is not as important as is implied in the literature,
specifically in the context of product launches by new
ventures.
Implications for Practice
Our findings provide fresh insights into how new ventures
in emerging industries can improve their new product
launch. We show that to introduce products successfully,
12
That market legitimacy matters more than scientific legiti-
macy is consistent with literature on new product development that
posits that market risk matters more than technical risk (Li and
Calantone 1998; Moorman 1995; Song and Montoya-Weiss 2001).
new ventures must engage in active legitimizing strategies.
These strategies provide new ventures with credibility in the
eyes of stakeholders and, in the process, bring them greater
rewards from product introduction. These rewards provide
much-needed revenues that, in turn, help new ventures sur-
vive and grow.
A key implication of our study for managers of new
ventures is that they need not and should not duplicate the
legitimacy gained through internal means with that gained
through alliances. The markets view such duplication as
unnecessary and costly. Thus, if new ventures do not have
internal legitimacy, they should form alliances. Conversely,
if they have internal legitimacy, they should bring their
products to market themselves. Moreover, we find that not
all forms of internal legitimacy are equal. A history of suc-
cessful product launches (historical legitimacy) has the
greatest impact on gains from product introduction, fol-
lowed by the presence of executives (market legitimacy)
and respected academics (scientific legitimacy) on the
board. Therefore, all else being equal, for firms that go
alone, it pays more to have introduced a new drug before,
followed by adding a reputed executive and then a reputed
scientist to the firm’s board.
12
Our analysis of product failures in addition to successes
shows that legitimacy is associated with lowered odds of
failure in the FDA approval process. These results further
highlight that investments in legitimacy help firms not only
reap greater financial rewards from their products but also
bring these products to market in the first place.
Marketers have extensively studied the positioning of
products. Our research suggests that new ventures in emerg-
ing industries also need to position themselves to gain more
from their new products. Given two products of the same
quality, the product introduced by the more legitimate firm
gains more. For new ventures facing intense competitive
pressures and serious resource constraints, these gains
could make the difference between life and death.
Limitations and Further Research
As with many early studies in an area, this research suffers
from several limitations, some of which offer fruitful
avenues for further research. First, we highlight four types
of internal legitimizing actions: historical, scientific, mar-
ket, and locational. These four dimensions may not be
exhaustive. Further research that examines additional
dimensions of legitimacy would be helpful. Second,
because of data limitations, we assume that all alliances
provide external legitimacy, and thus we do not examine
whether certain types of alliances with certain established
firms provide certain types of legitimacy or whether some
alliances provide greater external legitimacy than others.
Third, we examine the effects of legitimizing strategies on
new ventures that actually introduce new products. Further
research might study the effects of such strategies on new
The Fruits of Legitimacy / 73
13
We use the window (–1, +1) to be consistent with most prior
research. Nevertheless, we check for robustness by also doing the
analysis for the windows (0, 2), (–1, +2), and (–2, +2). Our results
remain qualitatively unchanged. To check for robustness, we also
create alternative dependent variables using period-adjusted and
market-adjusted models. The results using these variables are con-
sistent with the findings reported here and are available on request.
ventures that do not introduce products or those whose
products are not approved. Fourth, given our interest in
rewards to new ventures, we do not examine the returns to
the introduction of biotech products by established pharma-
ceutical firms. Rather, we focus on the stock market–based
returns to new products introduced by biotechnology firms.
It would be useful to examine whether the findings from
this industry generalize to other contexts and whether the
stock market effects we find hold for other performance
metrics, such as revenues and profits.
Appendix A: Measuring Change in
Market Capitalization
We measure ?MktCap
it
by employing a market-adjusted
model (Brown and Warner 1985). Thus:
where
?MktCap
it
= market capitalization change due to the
event at date t for stock i (FDA approval
date),
R
iT
= return on stock i at date T,
R
mT
= return on market index m at date T,
P
it – 2
= price of stock i at date t – 2, and
N
it – 2
= number of outstanding shares of stock i at
date t – 2.
13
First, our measure of change in market capitalization
calculates the absolute dollar value for rewards to product
introduction using abnormal returns. Thus, we follow other
researchers who have also attempted to compare across
many firms the gains (or losses) that accrue from an impor-
tant event (Dowdell, Govindaraj, and Jain 1992; Hendricks
and Singhal 1997; Sorescu, Chandy, and Prabhu 2003).
Although some empirical work using the event study
method employs percentage abnormal returns as the mea-
sure of interest, such work is typically concerned with
whether an event results in abnormal returns and not with
comparing the absolute gains that accrue across several
firms. In using a dollar measure, we are also consistent with
most research on the performance impact of marketing
actions. For example, research on advertising effectiveness
highlights the dollar revenues or dollar profits obtained as a
( ) A MktCap
R R
it
iT
Return on stock i
mT
Return o
1 ?
?

nnmarket index
Abnormal return


¸¸
¸

_
,

¸

1
]
1
1
1
?
+
?
T t
t
1
1
11
1
1
× ×
? ?
P N
it it 2 2
,
14
It might be argued that there is information leakage about the
new product before final FDA approval. For example, an FDA
advisory panel might approve the product, in which case the final
approval may be a mere formality. In such cases, relying only on
net present values (NPVs) generated at final approval might not
capture the “true” NPV associated with the product. To test for
this, we construct an alternative two-window dependent measure.
We do so by summing the NPV generated at the time of final FDA
approval and the NPV generated at the time of FDA advisory
panel approval, if any, for each product. We run regressions using
this alternative dependent variable and find that our results are
qualitatively unchanged. These results are available on request.
result of advertising and not whether these revenues had a
greater impact on the overall value of one firm versus
another based on percentage gains.
Second, by choosing the date of product introduction as
the event date, we can assess the impact of legitimacy on
the stock market gains to product introduction. Note that by
the time a product is introduced, the market is already
aware of the legitimacy variables (e.g., external, product
specific alliances) and has incorporated these into the valu-
ation of the firm as a whole. Therefore, if we were to find a
significant association between our measures of legitimacy
at the time of product introduction, all else being equal,
these associations would be a measure of how the valuation
of the particular product being introduced is affected by the
presence of various types of legitimacy.
Furthermore, whenever we refer to the date of “product
introduction,” we are referring to the FDA approval event.
The FDA approval virtually ensures that the new drug will
be launched in the near future, and thus the market reaction
to the FDA approval practically incorporates all the gains
associated with actual introduction. As such, this is the cor-
rect date to be used in event studies of returns to new prod-
uct introduction in the pharmaceutical industry (Sorescu,
Chandy, and Prabhu 2003).
In computing abnormal returns, it is crucial to ensure
that the event date is accurately noted. We obtain and cross-
check product introduction dates from multiple sources,
including FDA announcements, trade press tabulations, and
Dow Jones and Reuters wire services.
14
The ?MktCaps we
compute are reported in millions of dollars, adjusted using a
consumer price index deflator to reflect constant 1987
dollars.
Appendix B: Estimation Procedure
We rewrite Equation 2 for the firms that go alone in the fol-
lowing succinct form:
(B1) ?MktCap
i
= X
i
? + u
i
.
A firm’s decision to go alone is based on the underlying
selection Equation 4a, which can be summarized as
follows:
(B2) GoAlone
i
*
= Z
i
? + ?
i
,
where GoAlone
i
* is the net benefit of going alone and
GoAlone
i
= 1 iff GoAlone
i
*
> 0 and 0 if otherwise.
To estimate consistent parameters for Equations B1 and
B2, we use sample selection estimation methods (Heckman
74 / Journal of Marketing, July 2008
1979) that employ maximum likelihood. In the estimation
procedure, we account for the dichotomous nature of the
selection variable (GoAlone
i
= 1 for firms that go alone,
and GoAlone
i
= 0 for firms that form alliances) and jointly
estimate Equations B1 and B2. The dependent variable in
Equation B2 is “observed” for GoAlone
i
= 1 and “cen-
sored” for GoAlone
i
= 0. In other words, for GoAlone
i
= 1,
we observe Z
i
, X
i
, and ?MktCap
i
, but for GoAlone
i
= 0, we
observe only Z
i
and X
i
, and ?MktCap
i
is censored. We esti-
mate Equations B1 and B2 jointly using maximum likeli-
hood. We adopt a similar procedure for the alliance case. In
the case of firms that form an alliance, the formulation is
reversed, and the dependent variable is observed when firms
form an alliance and is censored when firms go alone. For a
discussion of the details of the likelihood function, its large-
sample properties, and its comparison to the two-stage pro-
cedure, see Maddala (1983).
REFERENCES
Aldrich, Howard E. and Ted Baker (2001), “Learning and Legiti-
macy: Entrepreneurial Responses to Constraints on the Emer-
gence of New Populations and Organizations,” in The Entre-
preneurship Dynamic, C. Schoonhoven, ed. Palo Alto, CA:
Stanford University Press, 207–235.
Angrist, Joshua (2001), “Estimation of Limited Dependent
Variable Models with Dummy Endogenous Regressors: Simple
Strategies for Empirical Practice,” Journal of Business and Sta-
tistics, 29 (1), 2–15.
Ashforth, Blake E. and Barrie W. Gibbs (1990), “The Double-
Edge of Organizational Legitimation,” Organization Science, 1
(2), 177–94.
Brown, Steven J. and Jerold B. Warner (1985), “Using Daily Stock
Returns: The Case of Event Studies,” Journal of Financial Eco-
nomics, 14 (1), 3–31.
Calantone, Roger J., Jeffrey B. Schmidt, and Michael Song
(1996), “Controllable Factors of New Product Success: A
Cross-National Comparison,” Marketing Science, 15 (4),
341–58.
Chaney, Paul K., Timothy M. Devinney, and Russell S. Winer
(1991), “The Impact of New Product Introductions on the Mar-
ket Value of Firms,” Journal of Business, 64 (4), 573–610.
Cialdini, Robert B. (1989), “Indirect Tactics of Image Manage-
ment: Beyond Basking,” in Impression Management in the
Organization: R.A. Giacalone and P. Rosenfeld, eds. Hillsdale,
NJ: Lawrence Erlbaum Associates, 45–56.
Cohen, Wesley M. and Daniel A. Levinthal (1990), “Absorptive
Capacity: A New Perspective on Learning and Innovation,”
Administrative Science Quarterly, 35 (1), 128–53.
Dacin, Tina M. (1997), “Isomorphism in Context: The Power and
Prescription of Institutional Norms,” Academy of Management
Journal, 40 (1), 46–81.
Das, T.K. and Bing-Sheng Teng (2000), “Instabilities of Strategic
Alliances: An Internal Tensions Perspective,” Organization Sci-
ence, 11 (1), 77–101.
Datar, Srikant and C. Clark Jordan (1997), “Advantages of Time-
Based New Product Development in a Fast-Cycle Industry,”
Journal of Marketing Research, 34 (February), 36–49.
Deeds, David, P. Mang, and M. Frandsen (2004), “The Influence
of Firms’ and Industries’ Legitimacy on the Flow of Capital
into High-Technology Ventures,” Strategic Organization, 2 (1),
4–25.
DiMaggio, Paul J. (1988), “Interest and Agency in Institutional
Theory,” in Institutional Patterns and Organizations, L.
Zucker, ed. Cambridge, MA: Ballinger, 3–22.
Dowdell, Thomas D., Suresh Govindaraj, and Prem C. Jain
(1992), “The Tylenol Incident, Ensuing Regulation, and Stock
Prices,” Journal of Financial and Quantitative Analysis, 27 (2),
283–301.
Dowell, Glenn and Anand Swaminathan (2000), “Racing and
Back-Pedaling into Future: New Product Introduction and
Organizational Mortality in US Bicycle Industry 1880-1918,”
Organization Studies, 21 (2), 405–431.
Dowling, John and Jeffrey Pfeffer (1975), “Organizational Legiti-
macy: Social Values and Organizational Behavior,” The Pacific
Sociological Review, 18 (1), 122–36.
Eisenhardt, Kathleen M. and Claudia Bird Schoonhoven (1996),
“Resource-Based View of Strategic Alliance Formation: Strate-
gic and Social Effects in Entrepreneurial Firms,” Organization
Science, 7 (2), 136–50.
Fama, Eugene F. (1970), “Efficient Capital Markets: A Review of
Theory and Empirical Work,” Journal of Finance, 25 (2),
383–417.
Ganesan, Shankar, Alan J. Malter, and Aric Rindfleisch (2005),
“Does Distance Still Matter? Geographic Proximity and New
Product Development,” Journal of Marketing, 69 (October),
44–60.
Gans, Joshua S. and Scott Stern (2003), “The Product Market and
the Market for ‘Ideas’: Commercialization Strategies for Tech-
nology Entrepreneurs,” Research Policy, 32 (2), 333–50.
Greene, William H. (2003), Econometric Analysis. Upper Saddle
River, NJ: Prentice Hall.
Grewal, Rajdeep and Ravi Dharwadkar (2002), “The Role of Insti-
tutional Environments in Marketing Channels,” Journal of
Marketing, 66 (July), 82–97.
Gulati, Ranjay (1998), “Alliances and Networks,” Strategic Man-
agement Journal, 19 (4), 293–317.
——— and M. Higgins (2003), “Which Ties Matter When? The
Contingent Effects of Inter-Organizational Partnerships on IPO
Success,” Strategic Management Journal, 24 (2), 127–44.
Hannan, Michael T. and John Freeman (1984), “Structural Inertia
and Organizational Change,” American Sociological Review,
49 (April), 149–64.
Heckman, James J. (1979), “Sample Selection Bias as a Specifica-
tion Error,” Econometrica, 47 (1), 153–61.
Hendricks, Kevin B. and Vinod R. Singhal (1997), “Delays in New
Product Introductions and the Market Value of the Firm: The
Consequences of Being Late to the Market,” Management Sci-
ence, 43 (4), 422–36.
Higgins, Monica C. and Ranjay Gulati (2003), “Getting Off to a
Good Start: The Effects of Upper Echelon Affiliations on
Underwriter Prestige,” Organization Science, 14 (3), 244–63.
——— and ——— (2006), “Stacking the Deck: The Effect of
Upper Echelon Affiliations for Entrepreneurial Firms,” Strate-
gic Management Journal, 25 (1), 1–25.
Klepper, Steven (2002), “The Capabilities of New Firms and the
Evolution of the US Automobile Industry,” Industrial and Cor-
porate Change, 11 (4), 645–66.
Lerner, Josh and Robert P. Merges (1998), “The Control of Tech-
nology Alliances: An Empirical Analysis of the Biotechnology
Industry,” Journal of Industrial Economics, 46 (2), 125–56.
Li, Tiger and Roger Calantone (1998), “The Impact of Market
Knowledge Competence on New Product Advantage: Concep-
tualization and Empirical Examination,” Journal of Marketing,
62 (October), 13–29.
Lunn, Mary and Don McNeil (1995), “Applying Cox Regressions
to Competing Risks,” Biometrics, 51 (2), 524–32.
The Fruits of Legitimacy / 75
Maddala, G.S. (1983), Limited-Dependent and Qualitative
Variables in Econometrics. Cambridge, UK: Cambridge Uni-
versity Press.
Mata, Jose and Pedro Portugal (1994), “Life Duration of New
Firms,” Journal of Industrial Economics, 42 (3), 227–45.
Matsuno, Ken, J. Mentzer, and Aysegul Ozsomer (2002), “The
Effects of Entrepreneurial Proclivity and Market Orientation on
Business Performance,” Journal of Marketing, 66 (July),
18–32.
McGarry, S., L. Rigg, M. Tracey, S. Brindle, and M. Readey
(2001), “Who Owns Biotech Drugs?” Goldman Sachs Global
Equity Research, Healthcare: Biotechnology (Europe) Report.
Moore, David S. and George P. McCabe (1998), Introduction to
the Practice of Statistics. New York: W.H. Freeman & Co.
Moorman, Christine (1995), “Organizational Market Information
Processes: Cultural Antecedents and New Product Outcomes,”
Journal of Marketing Research, 32 (August), 318–35.
Porter, Michael E. (1998), “Clusters and Competition: New Agen-
das for Companies, Governments, and Institutions,” Harvard
Business School Press Research Note.
Pouder, Richard and Caron H. St. John (1996), “Hot Spots and
Blind Spots,” Academy of Management Review, 21 (4),
1192–1225.
Prabhu, Jaideep C., Rajesh K. Chandy, and Mark Ellis (2005),
“The Impact of Acquisitions on Innovation: Poison Pill,
Placebo, or Tonic?” Journal of Marketing, 69 (January),
114–30.
Rao, Akshay R., Lu Qu, and Robert W. Ruekert (1999), “Signaling
Unobservable Product Quality Through a Brand Ally,” Journal
of Marketing Research, 36 (August), 258–68.
Ruef, Martin and W. Richard Scott (1998), “A Multidimensional
Model of Organizational Legitimacy: Hospital Survival in
Changing Institutional Environments,” Administrative Science
Quarterly, 43 (December), 877–904.
Satterthwaite, F.E. (1946), “An Approximate Distribution of Esti-
mates of Variance Components,” Biometrics Bulletin, 2 (6),
110–14.
Schmitz, James A. (1989), “Imitation, Entrepreneurship, and
Long-Run Growth,” Journal of Political Economy, 97 (3),
721–39.
Schumpeter, Joseph A. (1934), The Theory of Economic Develop-
ment. Cambridge, MA: Harvard University Press.
Shane, Scott A. (2001), “Technological Opportunities and New
Firm Creation,” Management Science, 47 (2), 205–220.
Shaver, J. Myles (1998), “Accounting for Endogeneity When
Assessing Strategy Performance: Does Entry Mode Choice
Affect FDI Survival?” Management Science, 44 (4), 571–85.
Song, Michael and Mitzi M. Montoya-Weiss (2001), “The Effect
of Perceived Technological Uncertainty on Japanese New
Product Development,” Academy of Management Journal, 44
(1), 61–80.
——— and Mark E. Parry (1997), “The Determinants of Japanese
New Product Success,” Journal of Marketing Research, 34
(February), 64–76.
Sorescu, Alina B., Rajesh K. Chandy, and Jaideep C. Prabhu
(2003), “Sources and Financial Consequences of Radical Inno-
vation: Insights from Pharmaceuticals,” Journal of Marketing,
67 (October), 82–102.
Srivastava, Rajendra K., Tasadduq A. Shervani, and Liam Fahey
(1998), “Market-Based Assets and Shareholder Value: A
Framework for Analysis,” Journal of Marketing, 62 (January),
2–18.
Stinchcombe, Arthur L. (1965), “Organizations and Social Struc-
ture,” in Handbook of Organizations, J.G. March, ed. Chicago:
Rand McNally, 142–93.
Stuart, Toby, Ha Hoang, and Ralph Hybels (1999), “Interorganiza-
tional Endorsements and the Performance of Entrepreneurial
Ventures,” Administrative Science Quarterly, 44 (June),
315–49.
Suchman, M.C. (1995), “Managing Legitimacy: Strategic and
Institutional Approaches,” Academy of Management Review,
20 (3), 571–610.
Swaminathan, Vanitha and Christine Moorman (2002), “Market-
ing Networks and Firm Performance,” working paper, Fuqua
School of Business, Duke University.
Van de Ven, Andrew, Roger Hudson, and Dean Schroeder (1984),
“Designing New Business Startups: Entrepreneurial, Organiza-
tional, and Ecological Considerations,” Journal of Manage-
ment, 10 (1), 87–107.
——— and Douglas Polley (1992), “Learning While Innovating,”
Organization Science, 3 (1), 92–116.
Zajac, Edward J. and James D. Westphal (2004), “The Social Con-
struction of Market Value: Institutionalization and Learning
Perspectives on Stock Market Reactions,” American Sociologi-
cal Review, 69 (3), 433–57.
Zimmerman, Monica A. and Gerald Zeitz (2002), “Beyond Sur-
vival: Achieving New Venture Growth by Building Legiti-
macy,” Academy of Management Review, 27 (3), 414–31.
Zucker, Lynne G., Michael R. Darby, and Marilynn B. Brewer
(1998), “Intellectual Human Capital and the Birth of US
Biotechnology Enterprises,” American Economic Review, 88
(1), 290–306.

doc_556441785.pdf