Content uploaded by M. Ann McFadyen
Author content
All content in this area was uploaded by M. Ann McFadyen on Nov 07, 2020
Content may be subject to copyright.
The Complementary Roles of Applied and Basic Research:
A Knowledge-Based Perspective
David H. Henard and M. Ann McFadyen
While a great deal of scholastic effort has gone into discovering the multifaceted
relationships between applied research initiatives and subsequent performance, rel-
atively little empirical research addresses the performance impact from firm invest-
ments in basic research initiatives. Even less addresses the interactive roles of both
types of research. The authors conceptualize and empirically evaluate the interac-
tive relationship between applied and basic research initiatives and firm perform-
ance. Applied and basic research projects are knowledge creation activities in a
product development domain, and both initiatives enhance the stored knowledge of a
firm. Stored knowledge is the fuel that drives the product development engine.
Applied research initiatives assimilate and exploit stored knowledge to develop new
products. Basic research initiatives contribute to and enhance the stock of knowl-
edge from which the applied initiatives are drawn. This expanded base of stored
knowledge has positive ramifications for subsequent applied research initiatives.
Results indicate that firms that engage in moderate or higher levels of applied re-
search will see enhanced performance returns from additional investments in basic
research. Conversely, firms that engage in relatively lower levels of applied research
see no performance enhancement at any level of investment in basic research. Firms
that rely on a flow of product innovations to provide a continued income stream must
certainly invest in applied research initiatives. However, additional investment in
directed basic research initiatives will augment future applied projects and could
become the source of sustainable competitive advantage.
Introduction
Innovation research initiatives can be broadly
classified as either applied or basic. While the is-
sue is debatable, there is widespread opinion that
firms have, on average, shifted the proportion of their
innovation research investments away from basic
research and toward applied research initiatives
(Cockburn and Henderson, 1998; Mansfield, 1980;
Rosenberg, 1990). Historically, applied research is
characterized as research that has fairly immediate
practical, and presumably profitable, ramifications.
Firm investments in applied research are considered
short-term, given that the ultimate end product is
marketed relatively quickly in relation to the resource
investment. By contrast, basic research is classically
portrayed as a more long-term investment in general
knowledge creation, having more unpredictable prac-
tical consequences.
There are differing views about the strategic value
of investing in applied and basic research initiatives.
One view is that firms should actively invest in applied
research for its anticipated performance gains but
should generally avoid direct investment in basic
The authors are listed alphabetically and thank the editor and two
anonymous referees for their insightful comments. We also thank
Deepak Sirdeshmukh and Sangkil Moon for their comments on ear-
lier versions of this research.
Address correspondence to: David H. Henard, North Carolina
State University, Raleigh, NC 27695-7229. Tel: (919) 515-8945. Fax:
(919) 515-6943. E-mail: david_henard@ncsu.edu.
J PROD INNOV MANAG 2005;22:503–514
r2005 Product Development & Management Association
research because knowledge contributions to the mar-
ketplace yield profit potentials too uncertain to di-
rectly justify the resource investment. Given the
competitive realities and the financial constraints of
firms, internal investments in basic research are often
not prescribed unless the firm has a clearly dominant
competitive position or the underlying marketplace
technology is inherently stable (e.g., Cassiman, Perez-
Castrillo, and Veugelers, 2002; Rosenberg, 1990).
However, another view is that focused investment in
basic research can yield products or technology that
can be profitably sold or licensed to others. Some
scholars argue that internal investments in basic re-
search are a crucial product development activity and
are a strong determinant of overall firm productivity
(e.g., Griliches, 1986; Mansfield, 1980). There are
some advocates for a mixed approach, calling for
actively pursuing simultaneous applied and basic
research agendas (e.g., Cockburn and Henderson,
1998; Nelson, 1959; Zucker, Darby, and Brewer, 1998).
The present research investigates the firm-level per-
formance returns that result from the interaction of
applied and basic research initiatives at 14 leading
manufacturing firms over a seven-year period. The use
of a unique data set
1
provides the ability to isolate
firm investments in both applied and basic research
initiatives. The present article contributes to the in-
novation literature by providing a conceptual frame-
work that helps explain why truly innovative firms
invest in both applied and basic research initiatives
and also by examining the moderating relationship
between the two research initiatives and the sub-
sequent return on investment (ROI). The following sec-
tion elaborates on the concepts that drive this investi-
gation. A description of the methodological approach
then is provided, along with the findings from the study.
The article concludes with a discussion of the results.
Conceptual Development
In keeping with the National Science Foundation
(NSF) classifications,
2
applied research initiatives are
defined here as research projects aimed at discoveries
that have specific commercial objectives. Basic re-
search initiatives are defined as research projects
aimed at the general advancement of knowledge that
do not have specific commercial objectives yet that
may be in fields of interest to the firm. Most compet-
itive companies either actively engage in product
development initiatives or seek to capitalize on the
investments of others. The rationale for investing in
applied research is self-evident. The development of
commercially successful products lies at the heart of
many firms’ quest for positive marketplace perform-
ance. Relative to basic research, the cycle time between
investment in applied research and the commensurate
return on investment is shorter. Hence, such invest-
ments fit well with the quarterly performance pres-
sures under which many firms operate. The rationale
for internal investment in basic research is a more
complex and, perhaps, more interesting question.
Increasingly, scholars view the product develop-
ment process through a knowledge utilization and
creation lens (e.g., Afuah, 2003; Day, 1994; Dickson,
1992; Kohli and Jaworski, 1990; Marsh and Stock,
2003; Moorman and Miner, 1997, 1998; Nelson, 1982;
Nonaka, 1991; van der Bij, Song, and Weggeman,
2003).
3
This growing perspective makes knowledge
creation theory (e.g., Cohen and Levinthal, 1990;
Nonaka, 1994; Simon, 1991) an appropriate concep-
tual framework for the present analysis. Knowledge is
created through a process by which information and
ideas are shared and combined through exchange
relationships (Nonaka, 1994; Simon, 1991). Applied
BIOGRAPHICAL SKETCHES
Dr. David H. Henard is assistant professor of marketing at North
Carolina State University. His research interests focus on corporate
identity/reputation, knowledge management, and new product de-
velopment issues. Dr. Henard’s work has been published in journals
such as the Journal of Marketing Research,Journal of the Academy
of Marketing Science,andPublic Relations Quarterly. Prior to his
career in academics, he worked 13 years for a large multinational
consumer products firm.
Dr. M. Ann McFadyen is assistant professor of management at
North Carolina State University. Her current research involves
studies that integrate knowledge creation and interpersonal network
theory as well as new product development. Dr. McFadyen’s work
is forthcoming in the Academy of Management Journal and Re-
search-Technology Management. Prior to her career in academics,
she worked 15 years for a large national bank in corporate treasury
new product development.
1
The Center for Innovation Management Studies, the Industrial
Research Institute, and the National Science Foundation provided
funding for the data collection.
2
The National Science Foundation classifications can be found at
http://www.nsf.gov/.
3
The related topics of knowledge, memory, and learning can alter-
natively be examined at various levels of a nested system—individual,
organizational, and social (March, 1991). This article approaches the
issue of firm knowledge utilization and creation within a product
development research domain from an organizational level.
504 J PROD INNOV MANAG
2005; 22:503–514
D. H. HENARD AND M. A. MCFADYEN
and basic research initiatives are prime examples of
knowledge creation activities in a product develop-
ment domain, and both initiatives enhance a firm’s
stored, or cumulative, knowledge. Over time when
independently constructed thoughts and ideas are
shared with others in the product development team,
common language, shared understandings, and intui-
tive insights are developed (Nonaka, 1994). Each of
these facilitates the creation of new knowledge and,
ultimately, new products. The knowledge gained from
such activities is an underlying mechanism by which
applied and basic research initiatives eventually lead to
performance returns.
The creation of new knowledge does not occur in
abstraction from current firm abilities (Kogut and
Zander, 1992). A tenet of knowledge creation theory
is that firms must possess stored knowledge in order
to generate new knowledge. Persistent investment in
knowledge acquisition (i.e., via research initiatives)
helps firms to better acquire, absorb, and assimilate
externally developed knowledge. This in turn makes
them better able to combine and exchange stored
knowledge to create new knowledge used in subse-
quent innovative endeavors (Cohen and Levinthal,
1990; Nonaka, 1994; Polanyi, 1966). In essence,
stored knowledge is the fuel that drives the product
development engine. Afuah (2003, p. 1) underscored
the importance of investment in knowledge by noting
that the path to innovation success is ‘‘strewn with the
bodies of firms that did not recognize the potential of
an innovation on time, did not have the right strategies
to exploit the innovation once they recognized it, [or]
did not know how best to implement the strategies. . . .’’
Investment in product development research di-
rectly impacts both the depth and breadth of the
firm’s stored knowledge. Because knowledge is built
cumulatively over time, previously stored knowledge
is the foundation upon which individuals will draw for
future innovations. Applied research initiatives assim-
ilate and exploit stored knowledge to develop new
products. Basic research initiatives contribute to and
enhance the stock of knowledge from which the
applied initiatives are drawn. For some firms, in-
vestments in basic research initiatives develop the
awareness of the latest technological advancements
in fundamental knowledge of a particular domain.
Other companies invest in partnerships with univer-
sities or other firms to enhance their level of stored
knowledge. This expanded base of stored knowledge
has positive ramifications for their subsequent applied
research initiatives. The ability to recognize, acquire,
and integrate information from external sources is
partially a function of stored knowledge; thus, re-
searchers must have some degree of common ground
to link new information with existing information.
This study argues that the ability of a company to
exploit knowledge is a critical component of its inno-
vative capabilities and that its ability to effectively
evaluate an array of currently available information is
largely a function of the depth and breadth of stored
knowledge. The depth and breadth dimensions of
stored knowledge are of particular importance to
product development activities. The depth of stored
knowledge reflects previous investments (i.e., time or
dollars) in acquiring related, relevant knowledge. A
central notion in the creation of a firm’s stored knowl-
edge is that the time required to accumulate the level
of this resource cannot be compressed. Knowledge
takes time to build, and companies that have histor-
ically possessed and created knowledge via persistent
research initiatives are better positioned to develop
new products than companies that have low initial
research activity. Basic research initiatives deepen the
firm’s stored knowledge base and enhance the depth
by increasing the awareness of the most recent tech-
nological advancements in a field, which may provide
the foundation for future applied research.
The breadth of stored knowledge is also important
to consider, as it provides the range of resources that
are drawn upon in applied research initiatives.
Diverse knowledge stimulates new ideas, promotes
creativity, and increases the researchers’ knowledge
integration skills (Utterback, 1971). By possessing a
broad base of knowledge, the firm has a diverse range
of knowledge to draw upon, thus better positioning its
product development researchers to make complex
and novel information linkages (Hamel and Prahalad,
1994). Breadth is obtained through diversity—the
broader the stored knowledge, the higher the proba-
bility the company’s employees will make unique link-
ages that result in successful innovations. Basic
research initiatives can enhance the breadth of a firm’s
stored knowledge if it is in fields that relate, at some
level, to the firm’s core activities or competencies (i.e.,
‘‘directed’’ basic research). Basic research initiatives
broaden the firm’s stored knowledge and therefore
enhance future applied research performance.
In order to fully benefit from either applied or basic
research initiatives, persistent investment appears to
be essential. Realistically, however, firm resources are
finite, and the two types of research initiatives com-
pete with one another for these scarce corporate
THE COMPLEMENTARY ROLES OF APPLIED AND BASIC RESEARCH J PROD INNOV MANAG
2005; 22:503–514
505
resources. March (1991, p. 71) stated that firms that
invest in basic research to the exclusion of applied re-
search ‘‘suffer the costs of experimentation without
gaining many of its benefits.’’ Conversely, March
(1991, p. 71) also noted that a reverse orientation is
likely to leave firms ‘‘trapped in suboptimal stable
equilibrium,’’ where research yields productive results
yet falls short of its potential because the groundwork
for future innovation is not fully being developed. He
therefore suggests that firms take a mixed research
approach. Unfortunately, research addressing the
interplay between applied and basic research invest-
ments is rare. Since the majority of empirical research
studies capture the relatively short-term performance
effects of product development investments, the
premise in this article is that the majority of existing
innovation research, unless explicitly noted otherwise,
evaluates applied research effects.
Managers and scholars generally agree that firm
investments in applied research initiatives are posi-
tively related to firm performance. The belief that
these activities positively impact performance is large-
ly supported by the voluminous investigations to
determine the optimal antecedents to new product
performance (see Henard and Szymanski, 2001;
Montoya-Weiss and Calantone, 1994). Investigations
of the impact of applied research on performance
have a long history in the innovation literature. The
SAPPHO studies (Rothwell et al., 1974) found five
factors that discriminate between performance success
and failure. Likewise, the NewProd (e.g., Cooper,
1979, 1980) and Stanford Innovation Project (Maid-
ique and Zirger, 1984) studies further developed
thought on the relationship between the applied re-
search efforts and subsequent performance returns.
Since these early studies, empirical investigations of
the relationship between applied research and the re-
sulting performance have increased dramatically. Yet
regardless of whether the investigations focused on a
product, process, or strategy dimension of applied
research, the results have consistently shown a posi-
tive and statistically significant empirical relationship
between applied research efforts and performance
returns (see Henard and Szymanski, 2001). Therefore,
we expect that
H1: The greater the firm’s level of applied research
investment, the greater the firm’s return on investment.
Whereas a positive relationship between applied re-
search initiatives and performance is well document-
ed, the relationship between basic research and
subsequent performance is relatively less investigated.
Rosenberg (1990) rhetorically asked why firms should
directly invest in internal basic research. The answer
to that question lies in one’s fundamental assumptions
about the goal of conducting basic research. If firm
management views the relationship between applied
and basic research as independent, then investment in
basic research may be difficult to conceptually or fi-
nancially justify except in instances where the firm
holds a clearly dominant marketplace presence. If,
however, basic research is viewed as complementing
the relationship between applied research and per-
formance returns, such investments may make more
sense. This complementary view posits that invest-
ments in basic research initiatives interact with invest-
ments in applied research initiatives. In effect,
knowledge creation theory suggests that investments
in basic research will impact the depth and breadth of
stored knowledge, which is essential to the assimila-
tion and exploitation of new knowledge in subsequent
applied research initiatives. This interaction positively
impacts a firm’s performance returns and is modeled
in Figure 1.
Cohen and Levinthal (1990) noted that the ability
of firms to exploit knowledge is a critical component
of its innovative capabilities and that its ability to ef-
fectively evaluate newly discovered information is
largely a function of its level of stored knowledge.
In addition to policies, procedures, and routines,
stored knowledge includes the basic skill sets of em-
ployees—some of which may be tacit (see Leonard-
Barton, 1992; Nelson and Winter, 1982; Polanyi,
1966)—as well as their understanding of past and
recent scientific or technological developments. In a
product development domain, the present study views
basic research as a prime contributor to the level of
stored knowledge, which is in line with the view of
Cassiman et al. (2002). This enhancement of the
Applied
Research
Initiatives
Basic
Research
Initiatives
H1:
H2: +
Returns
Figure 1. Complementary View of the Relationship between
Applied and Basic Research Initiatives on Performance Returns
506 J PROD INNOV MANAG
2005; 22:503–514
D. H. HENARD AND M. A. MCFADYEN
stored knowledge level of the firm bolsters research-
ers’ ability to exploit such knowledge in subsequent
applied research initiatives.
Scholarly investigations of the impact of basic re-
search investment on performance are relatively rare
(for exceptions, see Archibald and Finifter, 2003;
Averch, 1989; Griliches, 1986; Mansfield, 1980; Salt-
er and Martin, 2001) and are often concerned with the
societal effects of such research versus the firm level
effects of interest in the current research. Theoretical-
ly, by conducting basic research firms can more
effectively access the most recent technological ad-
vancements, which in turn increase the efficiency of
the firm’s own applied research (Cassiman et al., 2002;
Cockburn and Henderson, 1998). Thus, firms seeking
to enhance their applied research productivity and to
profit more efficiently from stored and newly availa-
ble knowledge should arguably maximize their com-
mitments to basic research. In line with Arrow (1962),
Rosenberg (1990, p. 170) views investments in basic
research as ‘‘a ticket of admission to an information
network.’’ He asserts that the output of basic research
is never a final product but rather is some form of new
knowledge that may be used to play some further role
in the development of new products (i.e., via applied
research initiatives). Conversely, firms that cease to
invest in knowledge may experience ‘‘lockout,’’ in
which their lack of persistent investment (e.g., in ba-
sic research) leads to a situation whereby the firm ei-
ther cannot assimilate and exploit newly developed
knowledge or simply cannot appreciate the value of
such knowledge when it becomes available (Afuah,
2003; Cohen and Levinthal, 1990).
The present study proposes that basic research does
not impact firm performance directly but rather en-
hances the impact of applied research on performance
returns. In essence, basic research efforts impact both
the depth and breadth of stored knowledge. Thus,
applied research efforts have an enhanced stored
knowledge base to draw upon, which should impact
firm performance more positively. Based on the
known literature, this interaction has not been empir-
ically tested. This research asserts that investments in
basic research initiatives moderate investments in
applied research initiatives to positively impact firm-
level performance returns. Hence,
H2: The greater the firm’s level of basic research in-
vestment, the greater the positive relationship be-
tween the level of applied research and return on
investment.
While the authors of this study advocate and test a
mixed research investment approach, the question of
degree remains. Griliches (1986) and Mansfield (1980)
empirically examined the impact of relative levels of
basic research on the productivity growth of firms
while controlling for the effects of other research and
development (R&D) investments. Their results, while
not directly applicable to the present study’s criterion
variable, provide some expectation that the greater
the level of investment in basic research, ceteris pari-
bus, the greater the expected performance returns.
The tenets of knowledge creation theory, with its em-
phasis on the power of combining stored and new
knowledge, simply reinforce this viewpoint. There-
fore, the relative association between applied and ba-
sic research investment and subsequent performance
also is examined. The following section details the
methodological approach taken to test the study’s
hypotheses.
Methodology
A multivariate longitudinal analysis was conducted by
sampling firm level measures from companies that are
members of the Industrial Research Institute (IRI), an
association comprised of companies considered lead-
ers in R&D. The source for firm-level financial data
was Standard & Poor’s (S&P) COMPUSTAT data-
base, which consists of over 35 years of time-series
financial data on approximately 7,300 active U.S.
companies, 5,000 inactive U.S. companies, 500 Cana-
dian companies, and 250 American Depository Re-
ceipts (foreign companies that file with the Securities
& Exchange Commission). The firm-level applied and
basic research measures were obtained through survey
data from a joint initiative between the IRI and the
Center for Innovation Management Studies (CIMS).
These data were collected over a seven-year time
frame (1992–1998) from 106 firms. The IRI/CIMS
annual surveys were designed to collect detailed data
pertaining to the sourcing and allocation of numerous
new product development activities from participat-
ing firms. The goal of the joint initiative was to build
a longitudinal, comprehensive database on product
development antecedents and outcomes.
The questions and definitions presented to partic-
ipants in the IRI/CIMS surveys remained constant
throughout the seven-year survey period. Over this
period, an average of 85% of the firms responded to
the survey each year. Firms with more than one year
THE COMPLEMENTARY ROLES OF APPLIED AND BASIC RESEARCH J PROD INNOV MANAG
2005; 22:503–514
507
of missing data were excluded from the database.
Firms that either responded to all years or had only
one year of partial data missing were retained in the
database. Any single-year missing values were re-
placed using a regression imputation procedure
(Little and Rubin, 1987). Regression imputation in-
corporated all reported values in a regression model
to determine an appropriate replacement value for
any missing data. Of the total participating IRI/CIMS
firms, 14 (13.2%) provided sufficient data over the
survey time frame. The final database contained seven
years of IRI/CIMS and COMPUSTAT data for 14
companies and included firms from consumer and
industrial SIC product groupings. The two firm-level
data sets were combined by company-provided
CUSIP numbers. Examination of key variables com-
mon to both the IRI/CIMS and COMPUSTAT dat-
abases indicated a greater than 90% correlation in
reported values across the two data sources.
Measures
Independent measures. Applied research initiatives
were defined as research projects aimed at discoveries
that have specific commercial objectives. To create an
appropriate and relative across-firm measure, the to-
tal dollars solely dedicated to applied research were
divided by the total dollars dedicated to R&D activ-
ities for each firm. Data for this variable were ob-
tained from the IRI/CIMS database. Basic research
investment was defined as research projects aimed at
the general advancement of knowledge. Participating
IRI/CIMS firms provided the total expenditures ded-
icated to basic research activities. To construct this
predictor variable and to create an empirically appro-
priate and relative measure across the firms in the
sample, the dollars dedicated to basic research were
divided by the total dollars dedicated to R&D activ-
ities for each firm. Data for this variable also were
obtained from the IRI/CIMS database.
Dependent measure. The financial performance
variable was operationalized as a firm’s ROI. Stand-
ard & Poor’s COMPUSTAT reports this measure as a
firm’s net income before taxes divided by its assets.
Return on investment is a firm-level financial per-
formance metric commonly used by managers to eval-
uate firm performance and has been used when
examining the impact of product development expen-
ditures on firm performance (e.g., Erickson and
Jacobson, 1992; Horwitch and Thietart, 1987). Data
for this variable were obtained from the S&P
COMPUSTAT database.
Control measures. Factors other than the explana-
tory variables might influence a firm’s ROI. There-
fore, the present model controlled for potential
extraneous effects. Overall firm investments in re-
search and development programs, as well as firm
size, may influence performance; therefore, to control
for any effects firm-wide expenditures in total research
and development activities might have on ROI, a firm
size-adjusted measure of R&D intensity was included
to capture the relative impact R&D investments
across firms has on subsequent firm performance.
Given that applied and basic research investments
are each a specific subset of total R&D resource in-
vestments, it is important to control for the perform-
ance impact of any other R&D activities (e.g., prod-
uct, process, technological, service) when interpreting
the relationships of interest in this study. Since extant
research (e.g., Rosen, 1991; Scherer, 1980) indicates
that firm size can influence R&D activities with larger
firms able to invest relatively more resources, a firm
size-adjusted measure is appropriate. To create an
appropriate and relative across firm measure and to
be consistent with existing R&D intensity measures
(e.g., Baysinger and Hoskisson, 1989; Coff, 2003; Ett-
lie, 1998; Helfat, 1994), total R&D expenditures were
divided by sales. Data for this variable were obtained
from the COMPUSTAT database.
Any potential heterogeneity of performance attrib-
utable to across-industry effects were controlled for
by including one dummy variable for the consumer and
industrial industry groupings represented by the re-
spondent firms. Six dummy variables also were includ-
ed for the seven years included in the regression to
control for any external environmental factors that
could have influenced return on investment (Helfat,
1994). Cognizant that performance in one year might
impact investments in subsequent years (Ben-Zion,
1984; Erickson and Jacobson, 1992; Grabowski and
Mueller, 1978; Helfat, 1994; Switzer, 1984), a panel data
method was used in this study to further control for any
effects due to serial correlation or heteroscedasticity.
Procedure
To ensure that these data did not contain a syste-
matic bias resulting from those firms retained in the
508 J PROD INNOV MANAG
2005; 22:503–514
D. H. HENARD AND M. A. MCFADYEN
database versus those eliminated, the responses of the
retained IRI/CIMS firms were compared with data
for the nonretained firms using both IRI/CIMS and
COMPUSTAT data. For these companies, gross
sales, number of employees, and R&D expenditures
were examined over the seven-year period. These were
pertinent variables commonly reported across both
databases. A Kolmogorov-Smirnov two-sample test
(Siegel and Castellan, 1988) was conducted to deter-
mine if any differences existed between retained and
nonretained firms in the distribution of responses for
the noted variables.
The large and statistically non-significant p-values
(p4.10) for each of the variables across the survey
period indicate that there is no statistical difference in
reported results between firms retained in the database
and those that were eliminated. Potential differences in
the distribution of firms responding across years were
further tested for. In each case examined, the p-values
for the variables exceeded .10, indicating no statistical
differences in the distributions and no sample variance
across years in the database. Given the results of these
tests, it can be concluded with confidence that both
the retained and nonretained firms are from similar
populations and that bias is not present.
The potential for multicollinearity is a concern
when using interaction terms as higher-order terms
tend to be highly correlated with the lower-order
terms of which they are comprised. To reduce con-
cerns of multicollinearity, Aiken, West, and Reno
(1991) recommended that all independent continuous
variables be standardized. Accordingly, all independ-
ent continuous variables were standardized to a mean
of zero and a standard deviation of 1. Standardization
serves two purposes in this study: It reduces multi-
collinearity among interaction terms and permits
graphical examination of the interaction terms by es-
timating coefficients for observations (1) one standard
deviation above and (2) one standard deviation below
the mean (Markham and McKee, 1991). Thus, high,
mean, and low investments in applied research can be
readily examined with high, mean, and low invest-
ments in basic research.
The data were examined for violations of assump-
tions of normality and multicollinearity. Because a
maximum likelihood methodology was used, statistics
such as variance inflation factors (VIF) were not avail-
able. The statistical software used in this research
(STATA 8SE) automatically checks for multicolline-
arity problems and will not calculate a model in the
presence of problematic collinearity. The software de-
tected no problems. As an additional measure, a man-
ual check was done for any multicollinearity before
fitting the model. Pair-wise correlations were examined
and were found to be low (all r.23). This, coupled
with an absence of sign flipping, leads to the conclu-
sion that multicollinearity is not a problem with these
data. The hypotheses were tested using a cross-sec-
tional time-series model using generalized least squares
adjusted for serial correlation and heteroscedasticity.
The conceptualization and empirical quantification
of firm investments in applied and basic research used
in this study imply that performance returns from ba-
sic research outputs are arguably more long-term than
those from applied research. Therefore, firm invest-
ments in applied research are captured contempora-
neously with that of ROI (i.e., both at time 5t) and
capture lagged investments in basic research (i.e.,
time 5t–1) to model the relationship more appropri-
ately.
4
This lagged structure in the model is important
because research indicates that firms that earn or ex-
pect a high ROI in one year have higher discretionary
funds and may increase new product development ex-
penditures in subsequent years (Erickson and Jacob-
son, 1992; Helfat, 1994). Other research indicates that
if managers do not expect a contemporaneous impact
on ROI from discretionary expenditures (e.g., R&D),
they will often focus more heavily on activities, such
as applied research, that lead to short-term gains
(Hayes and Abernathy, 1980). Thus, ROI from one
year understandably influences the resources available
for subsequent years’ applied or basic research invest-
ment. The construction of the present study’s varia-
bles and the time-series methodology used in this
research are employed to address the concern of re-
verse causality by examining the relationships over
a lagged time frame rather than simply contempo-
raneously.
Results
Table 1 provides the results of the hypotheses testing.
The base model, which includes only the control
variables as predictors, acts as a point of statistical
comparison for subsequent models. Positive and sta-
tistically significant coefficients in either model 1 or
the full model provides empirical validation for a hy-
pothesis. A statistically significant model difference
4
We thank an anonymous reviewer for this suggestion. Interviews
with four senior-level executives corroborate the construct validity of
using a lagged structure for basic research.
THE COMPLEMENTARY ROLES OF APPLIED AND BASIC RESEARCH J PROD INNOV MANAG
2005; 22:503–514
509
w
2
for model 1 or the full model indicates that the
respective model variables statistically predict firm-
level ROI after controlling for the base model varia-
bles. Support was found for H1—the greater the
firm’s level of applied research investment, the great-
er the firm’s return on investment (b50.011, p.01).
The full model incorporates firm investments in
applied and basic research as well as the hypothesized
interaction term. The moderation effect was tested
using criteria set forth by Baron and Kenny (1986).
First, H2—the greater a firm’s level of basic research
investment the greater the positive relationship be-
tween the level of applied research and ROI—is sup-
ported as the interaction between applied and basic
research is statistically significant (b5.011, p.05).
Second, it can be noted that basic research is uncor-
related with either applied research or return on in-
vestment (p4.10 for both). Third, the statistically
significant w
2
test of difference between the base
and full models (w
2
514.12, p.01) indicates that
investments in applied and basic research and the in-
teraction of the two have a statistically significant im-
pact on firm ROI over and above that of the control
variables. The w
2
test of difference between the full
model and model 1 is also significant (w
2
53.94,
p.05), which further substantiates H2. Descriptive
statistics for all modeled variables also are provided in
Table 2.
Previously, an interest was shown in assessing the
relative impact of basic research investments on the
relationship between applied research and ROI. To
calculate the nature of the interaction, we dismantled
the interaction effect by calculating the effects of ap-
plied research on ROI at low, mean, and high values
of basic research. Following Cohen and Cohen
(1983), we used Basic
L
, Basic
M
, and Basic
H
, which
correspond to low ( 1s), mean, and high ( þ1s)
levels of basic research investment, to generate a series
Table 1. Regression Results for the Relationship between Applied and Basic Research on Firm Performance Returns
Explanatory Variable Base Model
a
Model 1
b
Full Model
Intercept .059 (.01)
.048 (.01)
.046 (.01)
Industry Effect
c
—
—
—
Yearly Effect
d
——
—
R&D Intensity .001 (.00) .000 (.00) .001 (.00)
Applied Lagged Basic Effect — — .011 (.00)
Applied Research Effect — .011 (.00)
.014 (.00)
Lagged Basic Research Effect — .006 (.00)
.012 (.00)
No. of Firm Years 75 75 75
Model w
2
29.73
40.28
46.87
Wald w
2
Test of Difference between Models
e
11.00
14.12
a
Restricted model with only the control variables included. Standard Errors are in parentheses.
b
Restricted model with the control variables and the main effects of applied and basic research included.
c
Industry effects in the models are represented by dummy variables, which are omitted here for parsimony.
d
Yearly effects in the models are represented by dummy variables, which are omitted here for parsimony.
e
Represents the difference in explained variance between the Base model and the indicated model.
p.10.
p.05.
p.01.
Table 2. Descriptive Statistics and Intercorrelations for the Model Variables
Mean S.D. 1 2 3 4 5 6
Return on Investment .05 .05 —
Industry Effects .14 .34 .15 —
Yearly Effects .14 .35 .22
.02 —
R&D Intensity .37 .79 .02 .15 .02 —
Applied Research .24 .12 .23
.17 .01 .20
—
Lagged Basic Research .06 .07 .12 .06 .07 .07 .02 —
p.10.
p.05.
p.01.
510 J PROD INNOV MANAG
2005; 22:503–514
D. H. HENARD AND M. A. MCFADYEN
of regression equations of ROI on applied research
investment at specific values of basic research invest-
ment. This relationship is plotted in Figure 2.
The respective coefficients and standard errors
were computed to test the statistical significance of
each coefficient (Jaccard, Turrisi, and Wan, 1990).
T-test results indicated that the slopes for both mean
and high levels of basic research investment are dis-
tinct from zero ( p.01), while the slope for low
levels of basic research is not distinct from zero
(p4.10). Each additional unit invested in applied re-
search translates into a one-half standard deviation
increase in ROI (b5.02, p.01) when coupled with
mean levels of basic research investment. Likewise,
each additional unit invested in applied research pro-
duces an increase in ROI of nearly one standard de-
viation (b5.03, p.01) when coupled with high
levels of basic research investment.
Discussion
Summary and Contributions
While a great deal of scholastic effort has gone into
discovering the relationships between applied research
initiatives and subsequent performance, relatively
little empirical research addresses the performance
impact from firm investments in basic research initi-
atives. The present research substantiates established
empirical relationships from the plethora of applied
research initiatives and then advances the scholarly
knowledge base by conceptualizing and capturing the
complementary impact that firm investments in basic
research have on the relationship between applied re-
search initiatives and performance returns. Knowledge
creation theory suggests that firms draw upon stored
knowledge from basic research initiatives to create
new knowledge that enhances applied research efforts,
leading to innovation and product development. This
study’s results clearly demonstrate the moderating ef-
fect that basic research initiatives have on the rela-
tionship between applied research and firm-level
return on investment. Importantly, the present re-
search contributes to the literature by capturing and
empirically testing the performance impact that results
from investments in an intangible asset: knowledge.
Scholars increasingly view knowledge, and the abil-
ity to utilize knowledge, as one resource that provides
product development firms with the potential for en-
hanced firm performance (e.g., Cohen and Levinthal,
1990; Day, 1994; Henderson and Cockburn, 1994;
Kogut and Zander, 1996; Moorman and Miner, 1997;
Nelson, 1991; Nonaka, 1991; Penrose, 1959; Reed and
DeFillippi, 1990). Investments in applied and basic
research initiatives are the principal drivers of knowl-
edge creation in a product development domain,
which is why truly innovative firms invest in both ac-
tivities. Knowledge gained in one product develop-
ment area can be leveraged to enhance other areas.
For instance, knowledge gained from basic research
on the properties of polymer molecules in the late
1920s by DuPont scientists was leveraged 10 years
later to produce nylon and ultimately a host of other
synthetic fibers. Similarly, basic research by AMRAD
Corporation scientists in electrolytic capacitor chem-
istry and physics led to discovery of the transistor, as
well as to the theories and processes that became the
genesis of a global market estimated to be worth $6
billion.
Firms that rely on a flow of product innovations to
provide a continued income stream must certainly in-
vest in applied research initiatives. However, addi-
tional investment in directed basic research initiatives
will augment future applied projects and could be-
come the source of sustainable competitive advantage.
Those that can invest in internal basic research should
do so. Those that cannot should seek other methods
of leveraging basic research knowledge. For instance,
R&D employees should seek out the basic research
findings of others via journals, seminars, or work-
shops. Hewlett-Packard researchers regularly attend
academic workshops in search of new knowledge and
ideas. Partnerships with other firms or university
researchers, as well as hiring academically trained
engineers and scientists, also can help firms draw
upon research advances across a spectrum of disci-
plines. Indeed, evidence exists that the biotechnology
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
Low Mean High
Standardized Values of Applied Research Investments
ROI
Std. Low Basic
Std. Mean Basic
Std. High Basic
Figure 2. Interaction Plot of Applied and Basic Research
Investments
THE COMPLEMENTARY ROLES OF APPLIED AND BASIC RESEARCH J PROD INNOV MANAG
2005; 22:503–514
511
industry emerged as a result of private sector scientists
engaging in basic research with the academic
communities.
Firms that invest in low levels of applied research
will not realize performance differences by supple-
mental investments in any level of basic research.
Firms investing in mean levels of applied research
will see a positive impact on firm performance when
coupling the applied investment with either mean or
high levels of basic research investment. Firms that
invest in high levels of applied research will see an in-
substantial relative impact on performance when cou-
pling this investment with low levels of basic research,
will see an increased positive impact on performance
if coupled with mean levels of basic research, and will
see an even greater positive impact on performance
when coupled with high levels of basic research in-
vestment. Thus, even firms that invest relatively heav-
ily in applied research initiatives can further enhance
their performance returns with investments in basic
research, with the provision that the investment level
is at mean or higher levels.
This study’s results demonstrate that firm invest-
ments in basic research initiatives act in concert with
investments in applied research initiatives to impact
firm-level returns. Low levels of basic research invest-
ment ( 1s) do not enhance the impact of applied
research on firm performance. Firms that invest in
medium and high levels of applied research realize
enhanced performance when coupling these efforts
with mean levels of basic research investments. The
same relationship is found for high levels of basic re-
search investment ( þ1s). Finally, firms in this sample
that undertook relatively low levels of applied re-
search did not benefit from any level of basic research
investment. Thus, the line between applied research
and basic research is difficult to draw. While the di-
rection of applied research efforts is closely con-
strained by the practical problems that must be
solved, the direction of basic research may change
markedly as research proceeds and possibilities
appear (Nelson, 1959). The products from applied
research initiatives typically result from a series of
systematic actions. Knowledge gained from basic re-
search initiatives can interact with current and future
applied research efforts to develop novel or unique
products. Fundamentally, persistent investments in
basic research can enhance the depth and breadth of
a firm’s stored knowledge, which helps a firm facing a
number of potential applied research paths to choose
the path of greatest performance maximization.
Limitations and Directions for Further Research
While the inclusion of 14 companies from several in-
dustries and across seven years provides a good degree
of generalizability, a majority of companies in the
database are manufacturers of nondurable products.
While there is reason to believe that this study’s re-
sults also will hold for durable goods manufacturers,
the generalizability of these results is limited to man-
ufacturers of nondurable products. This data unique-
ly capture information from some of the most
innovative firms in the world. While this data are
not available from other sources and allow for a
unique examination of the research investment deci-
sions of clearly innovative firms, it does prevent the
collection of information on competing firms. Thus,
while relative scales were used to develop the model
variables, these variables do not represent relative
measures across all firms within a given industry.
The use of the COMPUSTAT database means that
all firms studied in this research are publicly traded
and, thus, are relatively large companies. Larger firms
have proportionately greater resources available to
devote to R&D activities than do smaller firms. Al-
though the authors see no conceptual distinctions, an
analysis of smaller companies might yield different
results due, in part, to resource constraints. Smaller
firms, or those with scarce resources, may focus more
heavily on applied research initiatives while acquiring
knowledge from basic research via outsourcing or al-
liance activities. Finally, evaluating the relationships
of interest would benefit from additional years of
observation. Even more longitudinal data—and the
subsequent increase in the number of available firm
years—would allow researchers to examine more
lagged period effects.
Further research in this area should focus on ex-
panding the base of firms to include not only a larger
number of firms but also firms that produce durable
and/or high-technology goods. An investigation of
the complementary relationship between applied and
basic research initiatives in a services domain would
also provide managers with greater insight and im-
prove managerial prescriptions. Likewise, an exami-
nation of the impact of alliances or outsourcing on the
relationship between applied and basic research and
performance returns is of contemporary interest, as is
the influence of merger and acquisition activity on the
relationships. This study only examined firms that in-
vested in both applied and basic research initiatives.
Incorporating findings from firms that invest in either
512 J PROD INNOV MANAG
2005; 22:503–514
D. H. HENARD AND M. A. MCFADYEN
applied or basic research to the exclusion of the other
and examining the associated performance differences
could improve future research. In summary, this study
provides strong evidence that the impact of applied
research investments on firm performance is enhanced
when coupled with investments in basic research. Im-
portantly, this impact occurs with average or above-
average investments in applied research and, at least,
average investments in basic research.
References
Afuah, Allan (2003). Innovation Management: Strategies, Implementa-
tion, and Profits, 2d ed. New York: Oxford University Press.
Aiken, Leona S., West, Stephen G. and Reno, Raymond R. (1991).
Multiple Regression: Testing and Interpreting Interactions. Newbury
Park, CA: Sage Publications.
Archibald, Robert B. and Finifter, David H. (2003). Evaluating the
NASA Small Business Innovation Research Program: Preliminary
Evidence of a Trade-Off between Commercialization and Basic
Research. Research Policy 32(4):605–19.
Arrow, Kenneth (1962). Economic Welfare and the Location of
Resources for Invention. In: The Rate and Direction of Inventive
Activity. National Bureau of Economic Research (ed.). Princeton:
Princeton University Press, 155–73.
Averch, Harvey A. (1989). Exploring the Cost Efficiency of Basic
Research Funding in Chemistry. Research Policy 18(3):165–72.
Baron, Robert A. and Kenny, D. (1986). The Moderator–Mediator
Variable Distinction in Social Psychological Research: Conceptual,
Strategic, and Statistical Considerations. Journal of Personality and
Social Psychology 51(6):1173–82.
Baysinger, Barry and Hoskisson, Robert E. (1989). Diversification
Strategy and R&D Intensity in Multiproduct Firms. Academy of
Management Journal 32(2):310–32.
Ben-Zion, Uri (1984). The R&D and Investment Decision and Its Re-
lationship to the Firm’s Market Value: Some Preliminary Results.
In: R&D, Patents, and Productivity. Zvi Griliches (ed.). Chicago:
University of Chicago Press, 229–312.
Cassiman, Bruno, Perez-Castrillo, David, and Veugelers, Reinhilde
(2002). Endogenizing Know-How Flows through the Nature of
R&D Investments. International Journal of Industrial Organization
20(6):775–99.
Cockburn, Ian and Henderson, Rebecca (1998). Absorptive Capacity,
Coauthoring Behavior, and the Organization of Research in Drug
Discovery. Journal of Industrial Economics 46(2):157–82.
Coff, Russell (2003). Bidding Wars over R&D-Intensive Firms:
Knowledge, Opportunism, and the Market for Corporate Control.
Academy of Management Journal 46(1):74–85.
Cohen, J. and Cohen, P. (1983). Applied Multiple Regression/Correla-
tion Analyses for the Behavioral Sciences. Hillsdale, NJ: Lawrence
Erlbaum.
Cohen, Wesley M. and Levinthal, Daniel A. (1990). Absorptive Ca-
pacity: A New Perspective on Learning and Innovation. Adminis-
trative Science Quarterly 35(1):128–52.
Cooper, Robert G. (1979). Identifying New Product Success: Project
Newprod. Industrial Marketing Management 8(2):124–35.
Cooper, Robert G. (1980). Project Newprod: Factors in New Product
Success. European Journal of Marketing 14(5–6):277–92.
Day, George S. (1994). The Capabilities of Market-Driven Organiza-
tions. Journal of Marketing 58:37–52 (October).
Dickson, Peter R. (1992). Toward a General Theory of Competitive
Rationality. Journal of Marketing 56:60–83 (January).
Erickson, Gary and Jacobson, Robert (1992). Gaining Comparative
Advantage through Discretionary Expenditures: The Returns to
R&D and Advertising. Management Science 38(9):1264–79.
Ettlie, John E. (1998). R&D and Global Manufacturing Performance.
Management Science 44(1):1–11.
Grabowski, Henry G. and Mueller, Dennis C. (1978). Industrial Re-
search and Development, Intangible Capital Stocks, and Firm
Profit Rates. Bell Journal of Economics 9:328–43 (Autumn).
Griliches, Zvi (1986). Productivity, R&D, and Basic Research at the
Firm Level in the 1970s. American Economic Review 76(1):141–54.
Hamel, Gary and Prahalad, C.K. (1994). Competing for the Future.
Harvard Business Review 72(4):122–28.
Hayes, R.H. and Abernathy, W.J. (1980). Managing Our Way to Eco-
nomic Decline. Harvard Business Review 58(4):67–77 (July–August).
Helfat, Constance E. (1994). Evolutionary Trajectories in Petroleum
Firm R&D. Management Science 40(12):1720–47.
Henard, David H. and Szymanski, David M. (2001). Why Some New
Products Are More Successful than Others. Journal of Marketing
Research 38:362–75 (August).
Henderson, Rebecca and Cockburn, Ian (1994). Measuring Compe-
tence? Exploring Firm Effects in Pharmaceutical Research. Strate-
gic Management Journal 15:29–44 (Special Issue).
Horwitch, Mel and Thietart, Raymond A. (1987). The Effect of Busi-
ness Interdependencies on Product R&D-Intensive Business Per-
formance. Management Science 33(2):178–97.
Jaccard, James, Turrisi, Robert and Wan, Choi K. (1990). Inter-
action Effects in Multiple Regression. Newbury Park, CA: Sage
Publications.
Kogut, Bruce and Zander, U. (1992). Knowledge of the Firm, Com-
binative Capabilities, and the Replication of Technology. Organi-
zation Science 3(3):383–97.
Kogut, Bruce and Zander, Udo (1996). What Firms Do? Coordination,
Identity, and Learning. Organization Science 7(5):502–18.
Kohli, Ajay K. and Jaworski, Bernard J. (1990). Market Orientation:
The Construct, Research Propositions, and Managerial Implica-
tions. Journal of Marketing 54:1–18 (April).
Leonard-Barton, Dorothy (1992). Core Capabilities and Core Rigidi-
ties: A Paradox in Managing New Product Development. Strategic
Management Journal 13:111–25 (Special Issue).
Little, Roderick J.A. and Rubin, Donald B. (1987). Statistical Analysis
with Missing Data. New York: Wiley.
Maidique, Modesto A. and Zirger, Billie Jo (1984). A Study of Success
and Failure in Product Innovation: The Case of the U.S. Electron-
ics Industry. IEEE Transactions on Engineering Management
31(4):192–203.
Mansfield, Edwin (1980). Basic Research and Productivity Increase in
Manufactuing. American Economic Review 70(5):863–73.
March, James G. (1991). Exploration and Exploitation in Organiza-
tional Learning. Organization Science 2(1):71–87.
Markham, Steven E. and McKee, Gail H. (1991). Declining Organi-
zational Size and Increasing Unemployment Rates: Predicting Em-
ployee Absenteeism from within- and between-Plant Perspectives.
Academy of Management Journal 34(4):952–65.
Marsh, Sarah J. and Stock, Gregory N. (2003). Building Dynamic
Capabilities in New Product Development through Intertemporal
Integration. Journal of Product Innovation Management 20(2):
136–48.
Montoya-Weiss, Mitzi M. and Calantone, Roger (1994). Determinants
of New Product Performance. Journal of Product Innovation Man-
agement 5(5):191–200.
Moorman, Christine and Miner, Anne S. (1997). The Impact of Or-
ganizational Memory on New Product Performance and Creativity.
Journal of Marketing Research 34:91–106 (February).
THE COMPLEMENTARY ROLES OF APPLIED AND BASIC RESEARCH J PROD INNOV MANAG
2005; 22:503–514
513
Moorman, Christine and Miner, Anne S. (1998). Organizational Im-
provisation and Organizational Memory. Academy of Management
Review 23(4):698–723.
Nelson, Richard R. (1959). The Simple Economics of Basic Scientific
Research. Journal of Political Economy 67(3):297–306.
Nelson, Richard R. (1982). The Role of Knowledge in R&D Efficiency.
Quarterly Journal of Economics 97(3):435–70.
Nelson, Richard R. (1991). Why Do Firms Differ, and How Does It
Matter? Strategic Management Journal 12:61–74 (Winter).
Nelson, Richard R. and Winter, Sidney G. (1982). An Evolutionary
Theory of Economic Change. Cambridge, MA: Belknap Press of
Harvard University Press.
Nonaka, Ikujiro (1991). The Knowledge-Creating Company. Harvard
Business Review 69(6):96–104.
Nonaka, Ikujiro (1994). A Dynamic Theory of Organizational Knowl-
edge Creation. Organization Science 5(1):14–37.
Penrose, Edith Tilton (1959). The Theory of the Growth of the Firm.
New York: Wiley.
Polanyi, Michael (1966). The Tacit Dimension. Garden City, NY:
Doubleday and Company, Inc.
Reed, Richard and DeFillippi, Robert J. (1990). Causal Ambiguity,
Barriers to Imitation, and Sustainable Competitive Advantage.
Academy of Management Review 15(1):88–102.
Rosen, Richard J. (1991). Research and Development with Asymmet-
ric Firm Sizes. RAND Journal of Economics 22(3):411–29.
Rosenberg, Nathan (1990). Why Do Firms Do Basic Research (with
Their Own Money)? Research Policy 19(2):165–74.
Rothwell, R., Freeman, C., Horsley, A., Jervis, V.T.P., Robertson,
A.B. and Townsend, J. (1974). Sappho Updated—Project Sappho,
Phase I. Research Policy 3(3):258–91.
Salter, Ammon J. and Martin, Ben R. (2001). The Economic Benefits
of Publicly Funded Basic Research: A Critical Review. Research
Policy 30(3):509–32.
Scherer, F.M. (1980). Industrial Market Structure and Economic
Performance, 2d ed. Chicago: Rand McNally College Publishing
Company.
Siegel, Sidney and Castellan, N. John, Jr. (1988). Nonpara-
metric Statistics for the Behavioral Sciences. New York: McGraw-
Hill.
Simon, Herbert A. (1991). Bounded Rationality and Organizational
Learning. Organization Science 2(1):125–34.
Switzer, L. (1984). The Determinants of Industrial R&D: A Funds
Flow Simultaneous Equations Approach. Review of Economics and
Statistics 66(1):163–68.
Utterback, James M. (1971). The Process of Technological
Innovation within the Firm. Academy of Management Journal
14(1):76–88.
van der Bij, Hans, Song, Michael and Weggeman, Mathieu (2003).
An Empirical Investigation into the Antecedents of Knowledge
Dissemination at the Strategic Business Unit Level. Journal of
Product Innovation Management 20(2):163–79.
Zucker, Lynne G., Darby, Michael R. and Brewer, Marilynn B.
(1998). Intellectual Human Capital and the Birth of U.S.
Biotechnology Enterprises. American Economic Review 88(1):
290–306.
514 J PROD INNOV MANAG
2005; 22:503–514
D. H. HENARD AND M. A. MCFADYEN
