Academy of Management Learning & Education,
2010, Vol. 9, No. 4, 638 – 651.
........................................................................................................................................................................
Does Business School Research
Add Economic Value
for Students?
JONATHAN P. O'BRIEN
Rensselaer Polytechnic Institute
PAUL L. DRNEVICH
The University of Alabama
T. RUSSELL CROOK
The University of Tennessee
CRAIG E. ARMSTRONG
The University of Alabama
The scholarly research conducted by business school faculty has long been the subject of
intense criticism for lacking relevance and value to practice. In contrast, we theorize that
such research is relevant and valuable in that it contributes to what is arguably the most
critical metric of relevance for business school students: the economic value they accrue
from their education. We investigate this counterargument on a sample of 658 business
schools over an 8-year period. We find that research adds significant value in that it can
potentially enhance student salaries by up to $24,000 per year. However, we also observe
that “excessive” research activity can lead to diminishing or even negative returns for
students, and a research focus solely on elite journals might rob students of the benefits
of exposure to a broader array of new ideas.
........................................................................................................................................................................
The scholarly research activity conducted at busi-
ness schools has come under harsh criticism as of
late with some business practitioners even argu-
ing that such research is a “vast wasteland” of
irrelevance (Bennis & O'Toole, 2005: 99). Yet this
criticism is not new. For almost 2 decades, scholars
themselves have been expressing concerns that an
excessive preoccupation with theory might bind
business school research into a “straightjacket”
that limits its relevance and value to practice (e.g.,
Bettis, 1991; Daft & Lewin, 1990). However, these
criticisms now seem to be reaching an almost fe-
verish pitch, with many prominent scholars sug-
gesting that research has overemphasized rigor
(e.g., following the scientific method) and theory
(Hambrick, 2007) at the expense of relevance and
value to practice (Bartunek, 2007; Hambrick, 2007;
McGrath, 2007; Pfeffer, 2007; Tsui, 2007). To make
matters worse, such research may be guiding what
faculty teach in the classroom (Rubin & Dierdorff,
2009), and according to Ghoshal (2005), may actu-
ally negatively influence practice.
Given these criticisms, in conjunction with the
centrality of research to most major business
schools (Rindova, Williamson, Petkova, & Sever,
2005), one might assume that graduate business
education is deeply troubled. Yet some evidence
suggests otherwise. Graduate business education
has experienced phenomenal growth over the last
3 decades, with over 100,000 MBA degrees awarded
in the United States in 2000 alone (Friga, Bettis, &
Sullivan, 2003; Morgeson & Nahrgang, 2008), and it
has remained popular with both students and
recruiters (Bradshaw, 2007). A recent survey by
the Graduate Management Admissions Council
638
Copyright of the Academy of Management, all rights reserved. Contents may not be copied, emailed, posted to a listserv, or otherwise transmitted without the copyright holder's
express written permission. Users may print, download or email articles for individual use only.
The authors wish to thank Ben Arbaugh, Neal Ashkanasy, Don
Bruce, Duane Ireland, Michelle Kacmar, Dave Ketchen, Barry
Mason, Bob Rubin, Donald Siegel, and several anonymous re-
viewers at the Academy of Management and the
Academy of
Management Learning & Education
for their comments, sugges-
tions, and contributions to the development of this article.

2010
O'Brien, Drnevich, Crook, and Armstrong
639
(GMAC) indicates that 94% of MBA graduates be-
lieve that their decision to pursue an MBA was the
“right” one (GMAC, 2006). Empirical evidence indi-
cates that the research agendas of business schol-
ars are indeed shaped by the problems that the
managers of large organizations consider impor-
tant (Barley, Meyer, & Gash, 1988). Further, in their
assessment of whether MBA curricula is aligned
with managerial competencies, Rubin and Dier-
dorff (2009: 22) found that “the relevant training
grounds [for managerial competencies] are likely
to be found in institutions of elevated research
activity rather than institutions where research is
de-emphasized and unsupported.” Recent research
has also provided empirical support for the posi-
tive relationship between academic research and
business school reputations (Rindova et al., 2005),
and rankings (Drnevich, Armstrong, Crook, &
Crook, In Press). Building on this evidence, we con-
tend that there is a vast gap between popular
perceptions and the reality of the relevance and
value of business school research (see Peng &
Dess, 2010, for a review of the discourse on this
issue).
We theorize that the scholarly research activity
conducted at business schools in general will con-
tribute to what is arguably the single most impor-
tant metric of relevance for students: the economic
value (i.e., salaries) they accrue from their educa-
tion. While recent studies have found some empir-
ical evidence of a relationship between research
and MBA salaries (Friga et al., 2003; Mitra &
Golder, 2008; Morgeson & Nahrgang, 2008; Rindova
et al., 2005), there remains a significant need for
additional research in this area. For example,
while Rindova et al. (2005) found that research has
a positive effect on the starting salaries of MBA
students, it is only an indirect one, mediated by the
reputation of the school. Mitra and Golder (2008)
also found that scholarly research has similar pos-
itive long-term effects on applicant, recruiter, and
academic perceptions as well as on performance.
However, the results of these two studies are not
surprising, given that both utilized measures
based on MBA starting salaries. Such measures
are problematic since new graduates are unknown
commodities, and it is reasonable to expect that
employers would base new hire decisions largely
on the reputation of the graduate student's institu-
tion (Rindova et al., 2005). In contrast, we utilize
measures based on average MBA salary appreci-
ation 3 years after graduation, which more accu-
rately reflects the value of the individual student's
knowledge, skills, and abilities. We also use a
dynamic panel data model that covers 658 gradu-
ate business schools from around the globe over
an 8-year period to investigate the relationship
between research and longer term economic value
for students. Using this approach also allows us to
more effectively control for a myriad of potentially
confounding factors that could induce a spurious
relationship between research activity and eco-
nomic value accrual for students.
While no observational study can definitively
establish causality, our approach and results do
provide strong evidence that the research con-
ducted by business school faculty does appear to
benefit the students of their schools by increasing
longer term salaries. Our evidence is in direct con-
trast to the claims of others who lament the lack of
relevance and value of academic research. How-
ever, we also observe a note of caution from our
findings in that the economic value created for
students appears to diminish, and can even turn
negative, when a business school's level of re-
search activity becomes excessive. Somewhat sim-
ilarly, we also observe that students benefit from
exposure to diversity in research activities, in that
some research published outside of the “A list jour-
nals” can also add significant value for students.
THEORY AND HYPOTHESIS DEVELOPMENT
Before we develop our arguments for the relation-
ship between the research conducted by business
school faculty and student economic value cre-
ation, we first acknowledge that even if we find no
relationship between research and value, this
finding would not necessarily invalidate business
school research. The Merriam-Webster (2008) dic-
tionary defines a university as “an institution of
higher learning providing facilities for teaching
and research . . .” Thus, even if we assume that
enhancing economic value for students is an im-
portant component of the
teaching
objectives of
business schools, the
research
function may still
exist as an independent institutional objective that
does not need to contribute to an institution's
teaching objectives. That is, true to the basic-
research mission of many universities (Scott, 2006),
the generation of new knowledge may be a vital
organizational objective, regardless of its immedi-
ate measurable economic impact.
1
Of course, even if research is not motivated by financial re-
turns, the knowledge emanating from it can lay the foundations
for advancements of enormous commercial significance (Blaug,
Chien, & Shuster, 2004), sometimes years or even decades after
the research is conducted (National Research Council, 2004).
1

640
Academy of Management Learning & Education
December
Confounding Considerations
Although pursuing scholarly academic research
may be a legitimate objective for business schools
even in the absence of explicit financial returns,
existing empirical evidence does suggest that re-
search and economic value creation for students
are at least correlated (e.g., Friga et al., 2003; Mitra
& Golder, 2008; Morgeson & Nahrgang, 2008;
Rindova et al., 2005). Casual observation would
also suggest that there is generally a positive cor-
relation between research and business school
rankings, even if the rankings do not explicitly
consider or effectively measure research activity
(Drnevich et al., In Press). Of course, such correla-
tions could well be spurious (i.e., noncausal) if
some unobserved other factor(s) was, in reality, the
true causal driver of both research activity and
student economic value creation (or MBA rank-
ings). While there are many possible factors that
could confound this relationship, we speculate
that a school's reputation, alumni network, and
financial resources may be the most likely candi-
dates. Of course, these potential confounding fac-
tors are not mutually exclusive.
First, a school's reputational capital may help it
to attract both faculty and students of higher qual-
ity (Rindova et al., 2005). Second, a strong alumni
network may help the school place its current stu-
dents in better positions, or with better employers,
while also allowing it to raise the funding neces-
sary to provide faculty both the time and resources
to pursue research. Third, independent of the
alumni network, a school's financial resources
may help it to offer students an enriched learning
environment while also allowing it to indulge fac-
ulty in their desire to conduct research. In the latter
two scenarios, the organizational resources de-
voted to faculty research might be considered a
type of perquisite that faculty consume when re-
sources are bountiful. Such an interpretation
would be consistent with Jensen's (1986) contention
that agency costs are most acute when financial
slack is abundant, or Bromiley's (1991) argument
that excessive organizational slack can lead to
inefficient resource allocation decisions.
Alternatively, faculty research activity may sim-
ply represent efficient resource allocation. Hosios
and Siow (2004) argued that faculty are the closest
thing that a university has to residual claimants
because increases in faculty salaries have been
driven largely by the residual that is left over after
accounting for all fixed costs and support staff.
Since faculty generally take an active role in the
governance of the university, if their research rep-
resented squandered resources, then we would ex-
pect to see many instances of the residual claim-
ants pressing for a more efficient resource
allocation model. For example, if faculty viewed
research as wasteful of university resources, we
would then expect them to seek to curtail new
hiring and increase teaching loads of existing fac-
ulty in return for commensurately greater pay.
However, since we observe little evidence of fac-
ulty pushing such initiatives, it is plausible that
faculty research activity may be a product of the
efficient distribution of the university's resources.
Although some may view the resources devoted to
research as a lavish perquisite, it may sometimes
be efficient for an organization to provide lavish
nonpecuniary benefits to employees (or managers)
if the employees will accept lower wages in return
for those benefits (Demsetz, 1983; Rajan & Wulf,
2004). Thus, if most faculty members enjoy doing
research (or at least enjoy it more than the other
components of their jobs, such as service or teach-
ing), then by offering faculty more time and re-
sources to conduct research, the university might
be able to recruit better faculty more cost effec-
tively. Hence, even if faculty research did not di-
rectly benefit students, students who attend
research-intensive universities might indirectly
reap more economic value than students who at-
tend teaching-focused ones.
The Positive Relationship Between Business
School Research and Economic Value
All of the above scenarios describe why there
might be a spurious relationship between faculty
research activity and student economic value cre-
ation. However, if we can demonstrate a relation-
ship between research activity and economic
value after controlling for the most relevant con-
founding factors, then it would provide strong
evidence that the research conducted by faculty
at a given institution does generally benefit the
students that attend that institution. It is our
contention that such a relationship does indeed
exist.
In order for research activity to generate eco-
nomic value for students, it must impart upon
those students valuable resources or capabili-
ties not shared by all MBA graduates (Barney,
1991). Developed modern economies are largely
knowledge-based economies, which indicates that
knowledge-based advantages are important for
high performance (Miller & Shamsie, 1996). Corre-
spondingly, faculty who are actively engaged in
research can likely provide value for their students
by transferring to them new knowledge gleaned
from their own research. In addition, even if an

2010
O'Brien, Drnevich, Crook, and Armstrong
641
individual faculty member's own research has lit-
tle relevance to practice, being actively engaged
in research helps faculty keep abreast of, and in-
volved with, “cutting edge” knowledge develop-
ments in the field. Faculty in turn can transfer such
knowledge to the student through classroom inter-
actions. Therefore, students are most likely to gain
knowledge-based advantages from faculty who
actively engage with a research community devel-
oping new intellectual capital and who stay
abreast of innovative developments in the field. In
contrast, such knowledge-based advantages are
unlikely to accrue to students at business schools
where faculty are not actively engaged in such
research communities.
Further, although rigorous, theory-driven re-
search has been criticized by some for being too
“scientific” (e.g., Bennis & O'Toole, 2005), we
think that faculty and students exposed to the
scientific process in such research might also
realize some benefits. In particular, active en-
gagement in knowledge creation through re-
search, as opposed to simply teaching from text-
books and educational materials that others
write, may help faculty hone their analytical
skills and, consequently, emphasize a more rig-
orous approach to problem solving and decision
making in the classroom.
2
Such rigorous problem
solving might resonate with students, and they
might make better decisions once they complete
their programs. Similarly, business school re-
search generally emphasizes the contingent na-
ture of relationships (e.g., Lawrence & Lorsch,
1967). As a result, students of research active
faculty will likely be better prepared to tackle
the complex problems in business today. As
Morgeson and Nahrgang (2008: 12) aptly observe,
“actively publishing faculty can take their
cutting-edge knowledge to the classroom to en-
hance the learning of their students, giving stu-
dents a competitive advantage compared to stu-
dents in programs where there is not as much
new knowledge.”
Collectively, these arguments suggest that stu-
dents should stand to benefit from the scholarly
2
research activities of their instructors. Thus, we
believe that business schools with higher levels of
research activity will create more economic value
for their students than business schools with low
levels of research activity. So long as labor mar-
kets are at least semicompetitive, graduates of
business schools with higher levels of research
activity should be able to appropriate at least
some of the economic rents attributable to their
knowledge, skills, and abilities (Coff, 1999). Stated
formally:
Hypothesis 1: There is a positive relationship be-
tween the level of a business
school's research activity and the
amount of economic value created
for students.
The Curvilinear Relationship Between Business
School Research and Economic Value
Although we believe that higher levels of re-
search activity will generally add economic
value for the students of a business school, we do
not contend that research is a continuous and
inexhaustible lever for improving economic
value for the students. In keeping with the law of
diminishing returns (Malthus, 1815), we would
expect that at some point, additional units of
research activity would provide negligible addi-
tional benefits to students. Hence, the relation-
ship between research activity and economic
value creation for students should be relatively
asymptotic. However, it seems likely that at some
point, the relationship would actually turn neg-
ative. Almost all organizations have limited re-
sources available to pursue their objectives (Bar-
ney, 1991), and the individuals within those
organizations have limited cognitive resources
to devote to multiple tasks (Norman & Bobrow,
1975). An excessive focus on one activity must, at
some point, come at the expense of other activi-
ties and objectives (Kaplan & Norton, 1992). Thus,
if a school places an excessive focus on research,
there might be less focus placed on teaching and
knowledge transfer. As a result, faculty will in-
vest relatively less effort in teaching and student
outcomes will suffer. Thus, while the research
activities of business school faculty might gen-
erally benefit students, excessive preoccupation
with research will likely attenuate this relation-
ship. Thus:
Hypothesis 2: There is a curvilinear (inverse U-
shaped) relationship between the
level of a business school's research
activity and the amount of economic
value created for students.
This sentiment was also echoed in a recent call for papers by
The UNESCO Forum on Higher Education, Research, and
Knowledge. One of the central themes of the colloquium was
that universities that lack the resources to sustain research
programs may be limited “to being institutions of knowledge
dissemination rather than knowledge creation” and that “[u]ni-
versities that are deprived, or deprive themselves, of that ingre-
dient risk the intellectual erosion of their programs of study,
lose their critical ability to assess claims to knowledge, and
become dependent on the outside supply of knowledge”
(UNESCO, 2006).

642
Academy of Management Learning & Education
December
METHODS
Sample
We had two major considerations for constructing
the sample used to test our hypotheses. First, be-
cause the MBA is an increasingly global product
(Bradshaw, 2007), we desired an internationally di-
verse sample of business schools. The second con-
sideration was a more practical one. Although
there are thousands of schools offering MBA-type
degrees in the world, we needed to limit the sam-
ple to a manageable number of schools for which
reasonably reliable data were available. Accord-
ingly, we chose to include in our sample all busi-
ness schools that are members of the Association
to Advance Collegiate Schools of Business
(AACSB) and participated in its Annual Business
School Questionnaire survey between the years
2001 and 2007. To construct the full sample (de-
scribed in detail below), we combined the AACSB
data with data from the
Financial Times
Global
MBA Rankings (henceforth FT Survey) for the years
2002 to 2009, and with publication count data de-
rived from the Institute for Scientific Informa-
tion's (ISI) Web of Knowledge database for the
years 1999 through 2006. Our final sample en-
compassed 658 different business schools and
3482 total observations.
lytical Considerations section). Finally, it is worth
noting that the salary data reported by the FT
Survey (and used in our analysis) are moving av-
erages, which take into account the previous 1–2
years of surveys to smooth out the effects of short-
term aberrations. This smoothing should not bias
our results in favor of our hypotheses, as it will
only reduce variability in our dependent variable,
making it more difficult to find statistical
significance.
Independent Variables
We measure the level of research activity of a
business school with the variable
publications,
which we constructed from data derived from the
ISI's Social Science Citation Index. We began by
downloading all citations for the years 1999 to 2006
for all journals listed under the subject areas of
“business,” “business, finance,” “management,”
and “operations research & management science.”
After excluding some nonacademic journals (e.g.
Forbes, Fortune),
our data encompassed 254 jour-
nals and 45,325 unique journal articles with 77,977
[nonunique] authors (i.e., an average of 1.7 authors
per paper). We then constructed the variable pub-
lications in three different ways. First, we con-
structed a measure that only included publications
in elite “A list journals” based upon their reputa-
tion for rigor and impact. According to Rindova
and colleagues (2005), the norms of modern science
prescribe that high-visibility publications such as
these serve as the most effective institutional cer-
tification of a faculty's research quality. Thus, we
used ISI's Journal Impact Factor ratings to develop
a list of the top-40 journals in business. Although
impact factor ratings are an imperfect measure of
journal quality, our approach is objective and our
derived list of journals corresponds closely to more
subjective journal lists used by the FT survey
and the University of Texas at Dallas' Top-100
Business School Research Rankings (see http://
top100.utdallas.edu/ for more information). Speci-
fically, we gathered data on the impact factor rat-
ings for the years 2003, 2005, and 2007 for the jour-
nals in our subject areas of interest. For each
journal, we averaged these three ratings, then
sorted the list and categorized the top-40 as “A
journals” (see Appendix). For every article appear-
ing in one of these 40 journals, we assigned a
fractional score based on the number of authors
(e.g., an article with one author is scored as a 1,
where an article with two authors is scored as 0.5,
etc.), then we computed a sum for every school
(based on author affiliations) for every year. Our
final measure for this level of research activity,
Variables
Dependent Variable
As we are interested in investigating whether the
research conducted by business school faculty
adds economic value for students, we measure this
value by way of the average percentage increase
in MBA
salary
(versus pre-MBA salary) 3 years after
graduation for the average alumni of a school.
Measuring percentage increase as opposed to sim-
ply measuring the raw salary 3 years after gradu-
ation helps to control for incoming student quality
and also allows us to assess how much the student
actually benefited from the MBA experience. Fur-
ther, long-term salary improvement is also argu-
ably one of the most important considerations for
business school applicants. We collected the data
for this variable directly from the
Financial Times
MBA rankings for the years 2002 to 2009. To the best
of our knowledge, the FT Survey provides the most
comprehensive and thorough information publicly
available on postgraduation MBA salaries. How-
ever, as this variable is only available for the
schools ranked by the FT Survey in any given year,
it is not available for every observation in our
sample (we address this issue below in the Ana-

2010
O'Brien, Drnevich, Crook, and Armstrong
643
A-publications,
is this sum divided by the total
number of full-time equivalent faculty members.
Aligning the
salary
data with the
publication
data presented some special considerations. The
FT Survey publishes its data early in the year
(generally January) from surveys conducted the
previous year of graduates from 3 years prior. Fur-
thermore, as the FT Survey uses moving averages
to smooth out aberrations, the salary data for a
given year may encompass students that actually
graduated 6 years prior. For example, the 2007 sal-
ary data consists of surveys done in 2006, mostly to
graduates from 2003, but possibly encompassing
graduates from 2002 and 2001. Further complicat-
ing matters is the fact that active and even rela-
tively mature research projects going on in (for
example) 2002 may not actually appear in journals
(i.e., be published) for several years. To account for
these lags, we matched salary data to the average
number of publications 3 and 4 years prior (e.g., we
merge salary data for 2007 with the average num-
ber of publications in 2003 and 2004). As a robust-
ness check, we also adjusted this lag structure a
year or two in each direction and found that doing
so does not materially alter our results.
As a further robustness check, we also examined
the possibility that research activity published in
journals outside of the “A journals” might also
create value for students. To examine this possi-
bility, we constructed the variable
B-publications
in the same manner as
A-publications.
Specifi-
cally, when we ranked journals based on their
average ISI Impact Factor, we classified all jour-
nals that ranked between number 41 and number
120 as B-journals (see Appendix). Finally, we also
constructed a third measure of research activity,
C-publications,
which encompassed all remaining
journals. In order to test the curvilinear (inverse
U-shaped) relationship we predict in Hypothesis 2,
we also include the square of each of the publica-
tion variables in each model.
Control Variables
In order to isolate the effect of the research con-
ducted by business school faculty on economic
value creation for students, we explicitly con-
trolled for factors that could induce a spurious
relationship between faculty research and student
outcomes. We use the variable
budget
to measure
the school's operating budget per full-time faculty
member to control for the financial resources of the
school. We believe that the operating budget is a
more pertinent measure of a university's financial
resources than alternative measures such as en-
dowment because many schools (especially non-
U.S. schools) are quite competitive despite having
little to no endowment. Further, in terms of both
research intensity and student outcomes, how
much the school is actually spending on its oper-
ations should be more important than the size of a
school's invested endowment. We also controlled
for the size of the
faculty
by including the number
of full-time equivalent (FTE) faculty at each school.
We further controlled for the
tuition
charged by the
school, since the cost of the education affects both
the school's finances and the overall value for
the money for the student. We constructed the
variables
budget, faculty,
and
tuition
from data
derived from the AACSB's DataDirect service
(AACSB, 2009).
We also control for a school's reputation
(Rindova et al., 2005; Safon, 2007) by constructing a
´
measure based on recent rankings. As the FT Sur-
vey did not rank many schools in our sample, but
points out that tiers are probably much more
meaningful than a school's absolute rank, we con-
structed dummy variables corresponding to rank-
ing tiers. Thus, we measure the variable
Tier 1
as
one if the FT Survey ranked the school in the top-50
within the last 2 years, and zero otherwise. Simi-
larly, we measure
Tier 2
as one if the FT Survey
ranked the school between 51 and 100 within the
last 2 years, and zero otherwise. A third tier, which
serves as the default condition, encompasses all
schools not ranked in the last 2 years by the FT
Survey. As a robustness check, we also experi-
mented with more tiers. Using more tiers did not
change our results, nor did it materially improve
model fit. Thus, we chose to present the more par-
simonious three-tier classification system.
We also included in the analysis a number of
additional dummy variables (U.S.,
Other-NA, Eu-
rope,
and
Asia)
to control for the geographic loca-
tion of the school. The default location would
hence be anywhere outside these areas (approxi-
mately 5% of the sample). Further, we also in-
cluded a dummy variable indicating whether the
school had a
doctoral
program. We derived data
for this item from the FT Survey and from the
AACSB data. In instances where this data item
was missing, we supplemented the data either by
performing web searches, contacting the school, or
by authors' judgments based on familiarity with
the schools.
Unfortunately, some constructs which we consid-
ered including, such as “quality of the alumni net-
work,” are difficult, if not impossible, to measure.
Fortunately, using a dynamic panel data model
allows us to introduce an additional level of im-
plicit control for potentially confounding factors
such as these (Wooldridge, 2003). To the extent that

644
Academy of Management Learning & Education
December
unmeasured factors such as “the quality of the
alumni network” are relatively stable over time,
including the lag of the dependent variable in the
model helps to control for the extent to which such
factors might affect the values of the dependent
variable,
salary
(see Wooldridge, 2005).
Analytical Considerations
Panel Data
Testing our hypotheses presented some critical an-
alytical considerations. First, we have multiple ob-
servations per school (i.e., panel data). With panel
data, if we fail to control for unobserved factors
that are correlated with both our dependent vari-
able and one or more of our independent variables,
then traditional ordinary least squares (OLS) re-
gression will produce biased results because the
afflicted independent variable(s) will be corre-
lated with the error term (Wooldridge, 2003).
Therefore, all models employ random effects for
the schools.
Censored Data
In addition to having multiple observations per
school, some data items were simply missing, as
the FT Survey did not rank every AACSB school in
our sample in a given year. This issue presents
another critical methodological concern because
the data are not just missing at random, but are
censored due to the school falling below a certain
threshold. In the FT Survey, the percentage in-
crease in
salary,
along with the highly correlated
gross weighted total salary, are by far the most
heavily weighted components of the rankings.
Thus, when the salary data are missing, it is safe
to assume that in general the school fared toward
the low end of the distribution on this item. There-
fore, estimating a conventional (e.g., OLS) model
on censored data will result in inaccurate esti-
mates because the sample is biased and not nec-
essarily representative of the full population
(Wooldridge, 2003). As a stylized example, consider
the data depicted in Panel A of Figure 1.
If we could measure all individuals in the hypo-
thetical population in Panel A, we would find a
correlation between the two variables of approxi-
mately 0.48. However, if we restricted our sample to
just high-performing observations (i.e., those scor-
ing above 70), we observe a correlation of only
0.026. As we illustrate in Panel B of Figure 1, a
censored sample might be even more problematic
if the true underlying relationship is nonlinear.
Despite this limitation, we can obtain valuable
FIGURE 1
Hypothetical Example: Censored Dependent
Variables
information from the observations with missing
values of the dependent variable if observations
that fall above and below the censoring threshold
have systematic differences in the distributions of
their independent variables. The two most com-
mon methods for making use of the information
inherent in the observations that fall below some
censoring threshold, and thereby correcting the
bias introduced by the censoring, are the Heckman
model and the Tobit model. The Heckman model is
most appropriate when the censoring is due to the
endogenous self-choice of the subjects, such as
when individuals decide not to work because the
available wage falls below their reservation wage
(Heckman, 1979). Conversely, the Tobit model is
more appropriate when the dependent variable
could potentially fall below the censoring thresh-
old (unlike hours worked, which cannot be nega-
tive), but we simply cannot observe it (Maddala,
1991). Since we expect that all schools would likely
prefer to rank in the FT Survey, and do in fact have
some [unobserved] score for salary, we deemed the
Tobit model to be more appropriate for our data.
We Winsorized the variable
salary
at the 1st per-

2010
O'Brien, Drnevich, Crook, and Armstrong
645
centile of its distribution and treated this value as
the censoring threshold.
RESULTS
We provide the descriptive statistics of our sample
in Table 1 and the results of our statistical analy-
ses in Table 2. Model 1 presents a standard (static)
cross-sectional random effects Tobit model. How-
ever, we only present this model for comparison
purposes and refrain from drawing inferences be-
cause it is static and does not include the lag of the
dependent variable, which can help serve as an
effective control for unobserved confounding
variables.
In Model 2, we present the dynamic Tobit model
that incorporates all the variables suggested by
Wooldridge (2005) to control for bias and unob-
served heterogeneity. For Model 2, we observe
highly significant Chi-square statistics (p
0.01),
indicating satisfactory model fit. As estimates for
time constant variables may be unreliable (Wool-
dridge, 2002: 541), we refrain from interpreting
those coefficients. However, the strong significant
coefficient on
salary (2002)
implies that there is
substantial correlation between the initial condi-
tion and unobserved heterogeneity (Wooldridge,
2005).
In regard to our hypotheses, the significant pos-
itive coefficient (p
0.01) for
A-publications
in
Model 2 indicates support for Hypothesis 1, that
there is a positive relationship between the level of
research activity and student economic value.
Likewise, the significant negative coefficient (p
0.01) for
A-publications
2
in Model 2 indicates sup-
port for Hypothesis 2, that the relationship between
the level of research activity and student economic
value is curvilinear (inverse-U). Taking the first
derivative of
salary
with respect to
publications
Endogeneity
Unobserved heterogeneity presents a problem
with our data because not only do we have multi-
ple observations per school, but also because
some third factor(s) could influence our prime the-
oretical variable,
publications,
as well as influ-
ence
salary.
Fortunately, using panel data pre-
sents the opportunity to help control for this
omitted variables bias. Hence, we address the po-
tential endogeneity issue by using the dynamic
Tobit model suggested by Wooldridge (2005). This
model allows us to implicitly control for the numer-
ous unobserved factors (e.g., alumni quality) that
might influence
salary
by including a lag of the
dependent variable as a predictor variable. Al-
though introducing the lag of the dependent vari-
able as a predictor variable can create bias, Wool-
dridge (2005) explains that a dynamic Tobit model
that effectively controls for unobserved heteroge-
neity can be estimated by simply including the
initial condition of the dependent variable in a
random effects model (i.e., the value in the first
time period), as well as some time invariant con-
trols, including a vector of all the values over time
of a time-varying variable. These variables will
absorb the correlation with the unobserved effects,
permitting accurate estimates of relationships
among the remaining variables. Hence, we include
as additional control variables the value of
salary
in 2002 and a vector of the values for
Tier 2
for each
school for the years 2003–2009.
TABLE 1
Descriptive Statistics
Variable
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
(12)
(13)
(14)
A-Pubs
B-Pubs
C-Pubs
Budget
Tuition
Faculty
Tier 1
Tier 2
U.S.
Other-NA
Europe
Asia
Doctoral
Salary
M
0.020
0.025
0.017
248.5
31.52
83.20
0.096
0.098
0.82
0.05
0.07
0.04
0.34
130.5
SD
0.04
0.04
0.02
370.4
36.11
53.85
0.29
0.30
0.39
0.21
0.25
0.19
0.47
31.2
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
(12)
(13)
0.74
0.72
0.18
0.20
0.11
0.46
0.25
0.05
0.01
0.05
0.02
0.34
0.03
0.80
0.22
0.20
0.08
0.27
0.22
0.08
0.00
0.02
0.07
0.32
0.15
0.19
0.17
0.07
0.25
0.22
0.11
0.06
0.02
0.07
0.28
0.16
0.71
0.15
0.17
0.10
0.37
0.01
0.05
0.63
0.12
0.05
0.17
0.18
0.11
0.24
0.06
0.00
0.52
0.14
0.01
0.26
0.23
0.30
0.09
0.18
0.09
0.42
0.02
0.19
0.05
0.02
0.16
0.05
0.36
0.32
0.06
0.05
0.05
0.01
0.34
0.38
3753.
0.47
0.56
0.43
0.09
0.24
0.06
0.04
0.07
0.08
0.05
0.08
0.20
0.03
0.03
0.06
Note:
For Salary, statistics are based on 578 observations. For all other variables,
n

646
Academy of Management Learning & Education
December
TABLE 2
Statistical Results of the Tobit Models
Model 1
Constant
A-Publications
A-Publications
2
B-Publications
B-Publications
2
C-Publications
C-Publications
2
Budget
Tuition
Faculty
Tier 1
Tier 2
U.S.
Other-NA
Europe
Asia
Doctoral
Lag-Salary
Salary-2002
Tier 2 - 2003
Tier 2 - 2004
Tier 2 - 2005
Tier 2 - 2006
Tier 2 - 2007
Tier 2 - 2008
Tier 2 - 2009
Observations
Wald Chi-square
Inflection Point
Max. Value
10.0
475.5**
1389.9**
0.03**
0.03
0.20**
60.47**
49.40**
15.66
13.71
32.27*
36.53*
16.23**
3753
859.8**
0.17
40.7
Model 2
24.1**
192.3**
610.9**
3.E 04*
9.E 04**
0.08
26.77
19.48
4.21**
11.21**
7.52
3.04**
6.91
0.23
0.24
2.00
6.54
2.39**
4.13
5.34
0.75
2.02
3482
1330.9**
0.16
15.1
Model 3
24.1**
335.1**
1845.7**
0.00
0.00
0.07**
26.95**
19.24**
7.88
15.20
1.19
13.20
6.78*
0.23**
0.24**
1.66
5.87
3.95
3.34
5.83
0.53
3.11
3482
1327.7**
0.09
15.2
Model 4
20.9*
234.2**
1189.8 
0.00
0.00
0.08**
25.80**
19.57**
2.87
11.70
5.76
7.02
6.88*
0.23**
0.26**
1.00
5.78
3.93
3.90
5.99
0.80
2.40
3482
1325.8**
0.10
11.5
Model 5
24.9**
137.5*
461.0*
264.3**
1597.6**
1.3
0.00
0.00
0.07**
26.51**
19.20**
8.73
16.00 
2.78
10.96
6.29*
0.23**
0.23**
2.07
6.60
3.19
3.66
5.11
1.08
2.36
3482
1314.7**
.15(A); .08(B)
10(A); 11(B)
All models also included year fixed effects (not reported for brevity).
 
p
0.10; *p 0.05; **p 0.01.
and solving for the inflection point reveals that the
maximum value on the curve occurs at 0.16 publi-
cations per FTE faculty member per year, which
would constitute approximately the 97th percentile
of the variable
A-publications
(the maximum value
was 0.44). Multiplying by our coefficients sug-
gests that this value for
publications
would add,
ceteris paribus,
about 15% to the average stu-
dent's salary.
Models 3 and 4 examine whether research activ-
ity that results in publications in lower tiers (e.g.,
“B” and “C”) journals also creates value for stu-
dents. Models 3 and 4 produce relatively similar
results to Model 2 using publications data based
solely on second- and third-tier journals, respec-
tively. In order to ascertain whether each tier of
publication adds value for students while control-
ling for other tiers of publications, we include all
three tiers in Model 5 (we do not include the square
of
C-publications
because it was insignificant and
including it increased potential collinearity prob-
lems). This model suggests that second-tier “B
journal” publications do indeed add significant
value over and above elite “A journal” publi-
cations, but third-tier “C journal” publications
generally fail to do so. Moreover, because of the
curvilinear relationships, a mix of A and B pub-
lications could potentially generate more value
than an exclusive focus on “A journal” publica-
tions. For example, student salaries maximize at
0.15
A-publications
and 0.08
B-publications
per
FTE faculty per year, which produces a predicted
annual increase in student salary of about 21%.
Conversely, 0.23
A-publications
and zero
B-
publications
produce a predicted increase of just
7.3%. This suggests that a moderately broad view
of academic scholarship, which can theoretically
encompass a wider array of new (and sometimes
even more radical) ideas, pays dividends to
students.
In order to facilitate the interpretation of our
results, we plot in Figure 2 the total contribution to
salary at varying levels of A- and B-level publica-
tions, as suggested by Model 5. We plot publica-

2010
O'Brien, Drnevich, Crook, and Armstrong
647
FIGURE 2
Impact of Faculty Research on Student Salaries.
The
x-axis
plots Publications (per FTE faculty per year)
from the 1st to the 99th percentile. (Note: the distribution of Publications varied slightly across journal tiers),
while the
y-axis
depicts the total increase in student salary (%) associated with the corresponding level
of publications. We derived predicted values from Model 5 of Table 2.
tions (per FTE faculty per year) from the 1st to the
99th percentile. However, publications per FTE fac-
ulty per year can be difficult to interpret, as many
schools employ numerous adjunct and clinical or
professionally qualified faculty, and may also
have numerous academically qualified faculty
who are no longer actively engaged in research.
Further, the distributions of publications are
highly skewed. Hence, in order to facilitate the
interpretation of our results, we use dotted lines to
designate the 95th to the 99th percentiles of publi-
cations. Intriguingly, this plot suggests that B-
publications can add comparable value to A-
publications, although the peak of their value
contribution occurs at a lower level than A-
publications, and they experience more sharply
declining returns to additional publications.
DISCUSSION
Our results indicate that the level of scholarly re-
search activity at business schools appears to add
considerable economic value to MBA students' fu-
ture salaries. Given that the average MBA salary
for programs ranked by the 2009 FT Survey is al-
most $115,000 per year, the 21% predicted increase
in salary that is possible at the optimal levels of
both types of publications would equate to a gain
of over $24,000 per year. This result strongly sug-
gests that research-intensive schools generally do
a superior job in helping their students acquire
and hone the knowledge, skills, and abilities,
which pays financial returns to the students
through their future employment. From these re-
sults, one might conclude that the actual state of
the relevance of business school research is not
nearly as dire as the common perceptions some
have suggested (e.g., Pfeffer & Fong, 2004;
Skapinker, 2008). However, we should also keep in
mind that an excessive focus on research activity
might start to erode that value premium, and that
the students of those schools appearing on the
diminishing returns side of the inverted-U curve
would likely benefit from lower levels of research
activity. Although excessive levels of research ac-
tivity attenuate student returns, more research-
intensive schools still generally produce better re-
turns for students than less research-intensive
schools. For example, even at the 99th percentile of
both
A-publications
and
B-publications,
student
salaries reap a predicted increase. Conversely, at
the 25th percentile of these distributions (which
equates to zero publications of each type), no value
is added for student salaries,
ceteris paribus.
The focus of our study was the general relation-
ship between the level of research activity in a
business school and student economic outcomes.
Accordingly, in our analysis we examined the re-
lationship between research publications and stu-
dent salaries, and considered both variation
across schools and within schools over time.
Hence, the results of our study have implications
for comparisons both across schools and longitu-
dinally within schools. To the extent that some
schools exhibit significant variation over time in
their level of research activity, our conjecture is

648
Academy of Management Learning & Education
December
that general trends are much more consequential
for student economic outcomes than year-to-year
fluctuations in publication counts. Indeed, al-
though post hoc analyses suggested there is more
variation in research activity across schools than
there is within schools over time, 16 schools did
exhibit a significant negative trend in publication
counts (using the total of A- and B-publications),
while over 100 schools exhibited a significant pos-
itive trend (p
0.05). In light of the illustrative
results of this simple post hoc analysis, it appears
an upward trend in research activity may likely be
beneficial for the students at most of the schools in
our sample, while a downward trend might actu-
ally be more beneficial for students at schools in
only the upper echelons of research activity.
Limitations, Implications, and Future Research
While our results have substantial practical impli-
cations for business schools and their stakehold-
ers, we do stress some important caveats. First,
while we demonstrate that scholarly research ac-
tivity generally adds economic value for students,
we cannot directly address whether this value is
attributable purely to rigorous, theory-driven
scholarly research. We can, however, likely as-
sume that our measure of
A-publications,
based on
the top-40 leading business journals, is reflective
of the same research priorities of the business
schools that have drawn so much criticism (e.g.,
Pfeffer & Fong, 2004; Skapinker, 2008). Second, we
also cannot directly address whether business
schools could have added even more value for
their students if they emphasized practical rele-
vance (in research and teaching) over rigorous,
theory-driven scholarly research; we can only ob-
serve that students appear to benefit economically
from business schools with more active scholarly
research programs. Third, while we conducted our
analysis on a large sample that is representative
of AACSB-accredited business schools, there are
thousands of schools offering MBA-type degrees,
and thus, our results may not be generalizable to
the entire population of business schools.
Future research could address some of these po-
tential limitations by extending our study in nu-
merous ways. First, it might prove enlightening to
examine whether the research activities of specific
business school disciplines add more value than
the research activities of other disciplines. As the
field of management tends to be much more inter-
ested in theory than some other disciplines (Ham-
brick, 2007), such an analysis might yield useful
fodder for the current debates regarding the theo-
retical straightjackets within management (Bar-
tunek, 2007; Hambrick, 2007; McGrath, 2007; Pfeffer,
2007; Tsui, 2007). Second, it might also prove useful
to differentiate the value added by normative ver-
sus positivist research activities. Speculatively, if
academically qualified business school faculty are
to add value in excess of what students may ac-
quire from professionally qualified academic staff
(e.g., clinical and adjunct faculty), then it may be
important to focus their research activities (like
engineering, law, and medical faculty), more on
how things should be done, as opposed to how
things are actually done in practice.
CONCLUSIONS
Our objective here was to investigate the apparent
gap between the popular negative perceptions of
the relevance and value of the rigorous theory-
driven research activity conducted at business
schools and the reality of the anecdotal (e.g., con-
tinued popularity and growth of business school
education) and empirical evidence to the contrary
(e.g., Friga et al., 2003; Mitra & Golder, 2008; Morge-
son & Nahrgang, 2008; Rindova et al., 2005; Peng &
Dess, 2010). To this end, we first theorized that such
research activity, in general, should likely contrib-
ute to what is arguably the single most important
metric of “relevance” for students: the economic
value they accrue from their education. We then
utilized a dynamic panel data model on a sample
of 658 business schools over an 8-year period to
empirically examine this relationship. Our study
offers both a response to some of the critiques of
business schools (e.g., Bartunek, 2007; Bettis, 1991;
Hambrick, 2007; Pfeffer & Fong, 2004; Pfeffer, 2007;
McGrath, 2007; Skapinker, 2008; Tsui, 2007), and a
validation and significant extension of some prior
related research (e.g., Drnevich et al., In Press;
Friga et al., 2003; Mitra & Golder, 2008; Morgeson &
Nahrgang, 2008; 2008; Rindova et al., 2005; Peng &
Dess, 2010). First, in response to the critiques and
commentaries, we find evidence that the research
conducted at business schools is relevant and
valuable to practitioners as evidenced by the con-
siderable longer term economic value added to
MBA student salaries. In terms of validation and
extension of prior research, we measured eco-
nomic value using salaries 3 years after gradua-
tion (as opposed to using MBA starting salaries),
which is more reflective of the value of the knowl-
edge, skills, and abilities provided by the business
school education. Additionally, we employed a dy-
namic panel data model, allowing us to implicitly
control for a myriad of potentially confounding fac-
tors for which we could not measure explicitly and
control that could induce spurious relationships in

2010
O'Brien, Drnevich, Crook, and Armstrong
649
studies using traditional models or cross-sectional
data. While only a controlled experiment with ran-
dom assignment can definitively establish causal-
ity, our work does provide strong evidence of the
probability of this causal link.
While our findings indicate that the scholarly
research activities of business school faculty do
indeed appear to benefit students economically by
significantly enhancing their longer term salary
appreciation, we also observe a note of caution.
The economic value created for students appears
to plateau, or can even diminish, when the level of
research activity becomes too excessive or is
overly restrictive. Thus, while more research activ-
ity in both “A” and “B” journals would appear to be
better for most schools, there appears to be dimin-
ishing returns from research that can even turn
negative from excessive levels of research activity.
These observations would appear to be useful for
deans and business school administrators in eval-
uating their schools' positions on research weight-
ing and in making decisions about resource allo-
cations and incentive systems. Finally, the results
of this study offer evidence to support the recent
conclusions of Ferris, Ketchen, and Buckley (2008:
743) that perhaps “we need to stop hand wringing
and apologizing for being organizational scien-
tists, and instead focus on pushing knowledge and
applications in this field forward in meaningful
ways.”
APPENDIX
Listing of ‘A' and ‘B' Journals Rank-Ordered by Average Impact Factor
Rank
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Tier
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
Journal
MIS QUART
ACAD MANAGE REV
ACAD MANAGE J
J MARKETING
MARKET SCI
J FINANC
J ACCOUNT ECON
ADMIN SCI QUART
J FINANC ECON
ORGAN SCI
STRATEGIC
MANAGE J
J MARKETING RES
J CONSUM RES
REV FINANC STUD
RES POLICY
J ACCOUNT RES
J MANAGE
MANAGE SCI
J OPER MANAG
HUM RESOUR
MANAGE
ORGAN STUD
J INT BUS STUD
INFORM & MANAGE
ACCOUNT REV
LEADERSHIP QUART
ORGAN RES
METHODS
J BUS VENTURING
ORGAN BEHAV HUM
DEC
TRANSPORT RES
B-METH
J MANAGE INFORM
SYST
J MANAGE STUD
J MONETARY ECON
MATH PROGRAM
ACAD MANAGE
EXEC
J PROD INNOVAT
MANAG
J ENVIRON ECON
MANAG
CORP GOV
SYST CONTROL
LETT
HARVARD BUS REV
J ACAD MARKET SCI
Avg. IF
4.538
4.347
3.520
3.498
3.217
3.056
2.918
2.784
2.699
2.497
2.483
2.164
2.161
2.084
1.817
1.758
1.729
1.689
1.688
1.654
1.651
1.642
1.641
1.625
1.609
1.591
1.524
1.516
1.506
1.499
1.452
1.431
1.421
1.385
1.375
1.375
1.374
1.374
1.366
1.329
Rank
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
Tier
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
Journal
J RISK UNCERTAINTY
J RETAILING
BRIT J MANAGE
INT J ELECTRON COMM
WORLD BANK ECON REV
J INF TECHNOL
J ORGAN BEHAV MANAGE
J PUBLIC POLICY MARK
EXPERT SYST APPL
INFORMS J COMPUT
DECISION SCI
DECIS SUPPORT SYST
MATH FINANC
OPER RES
ORGANIZATION
CALIF MANAGE REV
LONG RANGE PLANN
J FINANC QUANT ANAL
FINANC MANAGE
J BUS
INT J MANA G REV
INT J RES MARK
J QUAL TECHNOL
INT J SELECT ASSESS
MATH OPER RES
MIT SLOAN MANAGE REV
AM BUS LAW J
HUM RELAT
J CORP FINANC
J MONEY CREDIT BANK
INT J FORECASTING
J IND ECON
DISCRETE EVENT DYN S
TRANSPORT SCI
J MANAGE INQUIRY
J SCHEDULING
OMEGA-INT J MANAGE S
J ECON MANAGE STRAT
EUR J OPER RES
IEEE T ENG MANAGE
Avg. IF
1.327
1.309
1.291
1.287
1.268
1.263
1.228
1.185
1.160
1.143
1.127
1.127
1.124
1.119
1.092
1.068
1.067
1.064
1.060
1.018
1.017
1.014
1.008
0.976
0.976
0.960
0.957
0.933
0.925
0.921
0.920
0.913
0.884
0.877
0.876
0.851
0.844
0.843
0.842
0.836
Rank
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
Tier
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
Journal
RELIAB ENG SYST SAFE
ACCOUNT ORG SOC
ENTREP REGION DEV
TRANSPORT RES E-LOG
J WORLD BUS
NEW TECH WORK EMPLOY
COMPUT OPER RES
J FINANC INTERMED
QUEUEING SYST
J CONSUM AFF
J ADVERTISING
PSYCHOL MARKET
GROUP DECIS NEGOT
MANAGE LEARN
IND MARKET MANAG
TECHNOL FORECAST SOC
INT J OPER PROD MAN
NATL TAX J
J BUS RES
J GLOBAL OPTIM
J ADVERTISING RES
NETWORKS
TOURISM MANAGE
GROUP ORGAN MANAGE
REAL ESTATE ECON
WORLD ECON
FINANC ANAL J
INT J SERV IND MANAG
J INT MONEY FINANC
J OPTIMIZ THEORY APP
J BANK FINANC
BUS HIST
J INT MARKETING
IIE TRANS
INTERFACES
J OPER RES SOC
AUDITING-J PRACT TH
MARKET LETT
PROBAB ENG INFORM SC
J BUS ETHICS
Avg. IF
0.831
0.823
0.821
0.820
0.807
0.800
0.793
0.789
0.768
0.767
0.758
0.757
0.753
0.753
0.746
0.737
0.733
0.724
0.714
0.678
0.673
0.667
0.656
0.654
0.653
0.651
0.644
0.643
0.635
0.628
0.610
0.607
0.606
0.605
0.604
0.601
0.595
0.593
0.587
0.585

650
Academy of Management Learning & Education
December
REFERENCES
AACSB. 2009. The Association to Advance Collegiate Schools
of Business. http://www.aacsb.edu/knowledgeservices/
datadirect/
Barley, S. R., Meyer, G. W., & Gash, D. C. 1988. Cultures of
culture - Academics, practitioners and the pragmatics of
normative control.
Administrative Science Quarterly,
33(1):
24 – 60.
Barney, J. B. 1991. Firm resources and sustained competitive
advantage.
Journal of Management,
17: 99 –120.
Bartunek, J. M. 2007. Academic-practitioner collaboration need
not require joint or relevant research: Toward a relational
scholarship of integration.
Academy of Management Jour-
nal,
50(6): 1323–1333.
Bennis, W. G., & O'Toole, J. 2005. How business schools lost their
way.
Harvard Business Review,
83(5): 96 –104.
Bettis, R. 1991. Strategic management and the straightjacket: An
editorial essay.
Organization Science,
2: 315–319.
Blaug, S., Chien, C., & Shuster, M. J. 2004. Managing innovation:
University-industry partnerships and the licensing of the
Harvard mouse.
Nature Biotechnology,
22(6): 761–764.
Bradshaw, D. 2007. A new dawn for the MBA.
Financial Times,
January 29: 1.
Bromiley, P. 1991. Testing a causal model of corporate risk
taking and performance.
Academy of Management Journal,
34: 37–59.
Coff, R. W. 1999. When competitive advantage doesn't lead to
performance: The resource-based view and stakeholder
bargaining power.
Organization Science,
10(2): 119 –133.
Daft, R., & Lewin, A. 1990. Can organizational studies begin to
break out of the normal science straightjacket? An editorial
essay.
Organization Science,
1: 1–9.
Demsetz, H. 1983. The structure of ownership and the theory of
the firm.
Journal of Law and Economics,
26: 375–390.
Drnevich, P., Armstrong, C., Crook, T. A., & Crook, T. R. In
Press. Do Research and Education Matter to Business
School Rankings?
International Journal of Management
in Education.
Ferris, G. R., Ketchen, D. J., Jr., & Buckley, M. R. 2008. Making a
life in the organizational sciences: No one ever said it was
going to be easy.
Journal of Organizational Behavior,
29:
741–753.
Friga, P., Bettis, R., & Sullivan, R. 2003. Changes in graduate
management education and new business school strate-
gies for 21st century.
Academy of Management Learning &
Education,
2: 233–249.
Ghoshal, S. 2005. Bad management theories are destroying
good management practices.
Academy of Management
Learning & Education,
4: 75–91.
GMAC. 2006. MBA Alumni Perspectives Survey. Graduate Man-
agement Admission Council - Comprehensive Data Report.
McLean, VA: GMAC.
Hambrick, D. C. 2007. The field of management's devotion to
theory: Too much of a good thing?
Academy of Management
Journal,
50(6): 1346 –1352.
Heckman, J. 1979. Sample selection bias as a specification error.
Econometrica,
47: 153–161.
Hosios, A. J., & Siow, A. 2004. Unions without rents: The curious
economics of faculty unions.
Canadian Journal of Econom-
ics,
37: 28 –52.
Jensen, M. C. 1986. Agency costs of free cash flows, corporate
finance, and takeovers.
American Economic Review,
76:
323–329.
Kaplan, R. S., & Norton, D. P. 1992. Balanced scorecard: Trans-
lating strategy into action.
Harvard Business Review,
Jan –
Feb: 71– 80.
Lawrence, P. R., & Lorsch, J. W. 1967. Differentiation and inte-
gration in complex organizations.
Administrative Science
Quarterly,
12: 1– 47.
Maddala, G. S. 1991. A perspective on the use of limited-
dependent and qualitative variables models in accounting
research.
The Accounting Review,
66: 788 – 807.
Malthus, T. R. 1815. The grounds of an opinion on the policy of
restricting the importation of foreign corn.
The Works of
Thomas Robert Malthus,
E. A. Wrigley and D. Souden,
(eds.),1986, 7: 151–74. London: William Pickering.
McGrath, R. G. 2007. No longer a stepchild: How the manage-
ment field can come into its own.
Academy of Management
Journal,
50(6): 1365–1378.
Miller, D., & Shamsie, J. 1996. The resource-based view of the firm
in two environments: The Hollywood film studios from 1936 to
1965.
Academy of Management Journal,
39(3): 519 –543.
Mitra, D., & Golder, P. N. 2008. Does academic research help or
hurt MBA programs?
Journal of Marketing,
72(5): 50 – 68.
Morgeson, F. P., & Nahrgang, J. D. 2008. Same as it ever was:
Recognizing stability in the
Business Week rankings.
Academy of Management Learning & Education,
7:
26 – 41.
National Research Council. 2004.
Science, medicine, and ani-
mals.
Washington, DC: The National Academies Press.
Norman, D. A., & Bobrow, D. J. 1975. On data-limited and resource-
limited processes.
Cognitive Psychology,
7: 44 – 64.
Peng, M. W., & Dess, G. G. 2010. In the spirit of scholarship.
Academy of Management Learning & Education,
9: 282–
298.
Pfeffer, J., & Fong, C. T. 2004. The business school “business”:
Some lessons from the U.S. experience.
Journal of Manage-
ment Studies,
41: 1501–1520.
Pfeffer, J. 2007. A modest proposal: How we might change the
process and product of managerial research.
Academy of
Management Journal,
50(6): 1334 –1345.
Rajan, R. G., & Wulf, J. 2004. Are perks purely managerial ex-
cess? NBER Working Paper No. 10494.
Rindova, V. P., Williamson, I. O., Petkova, A. P., & Sever, J. M.
2005. Being good or being known: An empirical examina-
tion of the dimensions, antecedents, and consequences of
organizational reputation.
Academy of Management Jour-
nal,
48(6): 1033–1049.
Rubin, R. S., & Dierdorff, E. C. 2009. How relevant is the MBA?
Assessing the alignment of required curricula and required
managerial competencies.
Academy of Management
Learning & Education,
8(2).
Safon, V. 2007. Factors that influence recruiters' choice of B-
´
schools and their MBA graduates: Evidence and implica-
tions for B-schools.
Academy of Management Learning &
Education,
6: 217–233.

2010
O'Brien, Drnevich, Crook, and Armstrong
651
Scott, J. C. 2006. The mission of the university: Medieval to postmod-
ern transformations.
Journal of Higher Education,
77(1): 1–39.
Skapinker, M. 2008. Why business ignores the business schools.
Financial Times
[Asia Edition], January 8: 11.
Tsui, A. S. 2007. From homogenization to pluralism: Interna-
tional management research in the academy and be-
yond.
Academy of Management Journal,
50(6): 1353–1364.
UNESCO. 2006. Call for Papers: Universities as Centers of Re-
search & Knowledge Creation - An Endangered Species?
The UNESCO Forum on Higher Education, Research, &
Knowledge. Paris.
Wooldridge, J. M. 2002.
Econometric analysis of cross-sectional
and panel data.
Cambridge, MA: MIT Press.
Wooldridge, J. M. 2003.
Introductory econometrics.
Mason, OH:
South-Western.
Wooldridge, J. M. 2005. Simple solutions to the initial conditions
problem in dynamic, nonlinear panel data models with
unobserved heterogeneity.
Journal of Applied Economet-
rics,
20: 39 –54.
Jonathan P. O'Brien
is an assistant professor of strategic management at Rensselaer Poly-
technic Institute's Lally School of Management & Technology. He received his PhD in strategic
management from Purdue University. O'Brien's research interests include corporate gover-
nance, differences across institutional environments, and the strategic implications of a firms'
financial structure.
Paul Louis Drnevich
is an assistant professor of strategic management at the University of
Alabama. He received his PhD in strategic management from Purdue University. Drnevich's
research interests include competitive advantage and value creation/appropriation and the
effects of the dynamics of environmental uncertainty on performance, the implications of
capabilities and environmental factors for innovation and performance in entrepreneurial
ventures and small business, and the application of strategic management theory and meth-
ods to solving problems in management education.
T. Russell Crook
is an assistant professor of management at the University of Tennessee. He
received his PhD in strategic management from Florida State University. Crook's research
interests include strategy and entrepreneurship topics related to why some firms perform
better than others.
Craig E. Armstrong
is an assistant professor of entrepreneurship at the University of Alabama.
He received his PhD in strategic management from the University of Texas at San Antonio.
Armstrong's research interests include entrepreneurial decision making, bricolage, and re-
sourcefulness; entrepreneurship education, absorptive capacity, business models, and stra-
tegic competition.