for each year in our sample the number of unique authors publishing an article in the journals in our
sample in the previous five years. In 1988, for example, we calculated how many individuals published
an article between 1984 and 1988. This approach proxies for competition, because it assesses how many
individuals compete for journal space. We found a substantial increase in the number of people
publishing over time. In 1988, for example, we found that 1572 individuals published an article between
1984 and 1998. In the last year of our sample, 2008, we found that this number increased to 3158. This
is important information, as we are now better able to discuss changes over time. With this addition we
do not emphasize demographic shifts until we reach our discussion. We hope that our revised discussion
now more clearly develops the linkages to which you refer.
5. COMMENT: p. 3 - competing for space - I agree with the way you portray this context but would like
to see you more completely describe it. For instance, it seems that the growth in number of competing
authors needs to be mentioned here, as I don't think you get to this until much later in the paper. In
fact, I'd like to see you get this paper to the point where you clearly and cleanly can contrast the level
and nature of competition across the different periods of time. Perhaps here is where you start with
something about number of authors (though this is different than potential authors) and the level of
concentration or fragmentation (i.e., average papers per author and number of authors with more than
5 in 5 and 10 in 10). Then after you test your hypotheses you can present a side by side comparison of
how this has changed (along the three dimensions).
RESPONSE: Thank you for this great suggestion. As we discussed in our response to your previous point,
we do just this. In the revision we introduce Figure 2, which illustrates how competition has evolved
over time in management as well as the subdisciplines. This new figure will allow readers to compare
the figures in Tables 2a and 2b with the increasing competition. We might speculate, then, that
achieving these milestones has become more difficult over time.
6. COMMENT: p. 7 -- underlying process -- its about H2 that it strikes me that you might help the reader
understand whether this is (1) simply a numbers game -- i.e., more competitors -- or (2) a change in the
nature of competition from broad-brush to incremental research contributions. When our field was
young, and there were relatively few scholars, in many way the field was more munificent in terms of
research opportunities. Today, the field is much more topically fragmented, and most new studies offer
only incremental, albeit highly sophisticated, advances. Replications too are not welcome in top
journals, such that these later authors have to do new, incrementally novel things. The general context
in which researchers find themselves seem to lend itself less to a few high-volume producers and more
to a situation favoring many authors an topic and methodologically fragmented world. This AMR might
be useful in thinking about how theory development and testing has evolved and thus contributed to
the changing competitive landscape that you study.
Fabian, F.H., 2000. Keeping the Tension: Pressures to Keep the Controversy in the Management
Discipline. Academy of Management Review, 24 (2): 350-371.
RESPONSE: Thank for you providing these ideas. We now explicitly cite Fabian’s (2000) work and discuss
more scholars trying to compete (numbers game) as well as increased sophistication and fragmentation
(intensity of incremental research). We think that this increases the clarity of the underlining processes
directing the conclusions proposed in H2. We hope that you agree.
7. COMMENT: p. 7 -- "increased munificence may moderate the negative effects" – this reads like a
hypothesis where there is an explicit interaction. I think you want to reword this so that it better maps
to your main effect prediction. Besides, I'm not what the functional form of this would look like.
RESPONSE: Thank you for this point. In the revision, we have changed this wording.
8. COMMENT: p. 7 -- micro/macro -- this makes sense and is interesting but somehow I'd like to see you
better foreshadow how you eventually test this. You also talk about some of the fundamental difference
between micro and macro publishing but don't talk about the fact that micro research has a higher
number of co-authored papers (at least I assume it is a fact and your data can tell me if that is true). This
difference in coauthorship would seem to have important implications for your results as well, since a
paper with five authors would appear many more times than a paper with one or two authors (apologies
if you dealt with this key issue but I missed it).
RESPONSE: You raise an important point with respect to coauthorship over time. This was also a point
raised by Reviewer 2. In our original submission, we tried to implicitly note this by distinguishing
between each journal’s articles and authoring events.
To more explicitly examine this issue, we now include Figure 3, which shows how coauthorship has
changed over time for each sub-discipline. Thank you for this valuable suggestion.
9. COMMENT: p. 10 -- micro/macro coding -- I think this is a great section and I'd like to see you
foreshadow this stuff a bit starting on p. 7. When you talk about SMJ (one of my home journals) you
might add a sentence that states: While SMJ does publish micro papers, any such papers that were not
counted based on our heuristic would have made the micro results that much stronger. Or something
like that. Your coding may slightly under-represent micro papers, but your results would only be
stronger if you coded each paper in SMJ.
RESPONSE: This is an interesting perspective. We have added language suggesting that our coding
scheme is not perfect. Our sense is that some micro researchers might argue that some of articles
published in micro journals examine macro topics as well. At the end of the day, we are not sure that
one group has an advantage in this respect. As such, we now include this issue as a limitation of our
10. COMMENT: p. 10 -- tautology? -- Can you help the reader better understand why, if you sampled
more journals with micro papers why you would not find more micro authors? And wouldn't this
problem be amplified in your data if there were more micro authors per paper? If this were a regression,
its almost like I'd like to see you be able to say we controlled for the number of predominately micro
and macro journals and still found that micro authors...
RESPONSE: This is another good point. In the revision, we attempt to deal with this issue. First, we add
Organization Science to our set of journals. Second, we try to do a better job highlighting the fact that
our results are substantively similar when we use JAP as our only micro journal. Relative to OBHDP and
PP, JAP publishes a larger number of articles.
11. COMMENT: p. 11 -- "this procedure revealed..." -- I could not tell from this paragraph and later in the
paper if you included all these papers in your analyses, or excluded most of the 840 articles/authors.
RESPONSE: These authors were utilized where appropriate. We could not determine their research foci,
and the vast majority did not publish 5 articles in our sample journals. Thus, they were less useful for our
analyses that examined differences between the sub-disciplines. However, we did include such authors
when examining the management field as a whole (e.g., Figure 1). We hope that this makes our
methodology more clear.
12. COMMENT: p. 14 -- middle paragraph following "insert Table 2" -- reiterate my question or concern
about number of authors for micro papers. Across the sets of authors, how many actual unique papers
were there? I know this won't be the case, but what if 177 = 5 authors on the same 35 papers and 76 is 2
authors on 38 papers -- it won't be the case but I hope you get my meaning. How many unique papers
are there? It seems that competition is at the level of the paper, and if authorship varies systematically
across micro and macro, then you aren't communicating or comparing apples to apples.
RESPONSE: This is another great point. As we indicated previously, we now explicitly document
coauthorship practices with Figure 3. We think this information illustrates quite clearly that
coauthorship varies by sub-discipline. Additionally, as we mentioned in our response to Comment # 4,
we have added new analyses that investigates unique authors. Specifically, we calculated for each year
in our sample the number of unique authors publishing an article in the journals in our sample in the
previous five years. In 1988, for example, we calculated how many individuals published an article
between 1984 and 1988. This measure proxies for competition, because it assesses how many
individuals are competing for journal space. We found a substantial increase in the number of people
publishing over time. In 1988, for example, we found that 1572 individuals published an article between
1984 and 1998. In the last year of our sample, 2008, we found that this number increased to 3158.
Again, we appreciate all of your insightful comments. We worked hard to be responsive to them. Thank
you for taking the time and energy to help us improve the paper.
RESPONSES TO REFEREE #2’s COMMENTS:
1. COMMENT: I list my specific concerns for the manuscript below.
RESPONSE: Thank you for all of your detailed comments and suggestions. We found them quite useful as
we approached our revision.
2. COMMENT: Given the journals selected (i.e. mostly micro in nature, it seems nearly impossible that
H4 would not be supported given that JAP, OBHDP, and Psych are all exclusively micro.
RESPONSE: This is a good point that Reviewer 1 also pointed out. In the revision we attempted to
address this issue in two ways. First, we expanded the number of journals in our sample to include
Organization Science. Second, we attempt to underscore that we ran supplementary analyses where we
included only one journal as our primary micro journal, JAP, and the results were substantively similar.
Looking at the dispersion of journals over time, JAP represents a large majority of the articles published.
So, we conclude it is not a merely our sample selection that matters, but instead our finding is
representative of trends in the field.
3. COMMENT: I think an interesting, but unexplored point is whether academics have changed their
research strategy in light of the increased expectations of research production. One strategy in
particular, using more co-authors is a way in which researchers can increase their productivity, even if
space in top journals is limited. This strategy may be particularly salient to macro researchers, given the
extremely limited space available.
RESPONSE: Thank you for this valuable comment, which was echoed by Reviewer 1. Based on this
suggestion, we now include in our discussion section more discussion of coauthorship over time. In
addition, we now created Figure 3, which illustrates how coauthorship for each journal evolved between
1988 and 2008. As seen in Figure 3, there is a slightly positive trend in terms of coauthoring. Thank you
again for this valuable suggestion.
4. COMMENT: I found interpreting the results of the empirics used to test the hypotheses nearly
impossible due to the fact that no tables or correlations or regression coefficients were offered. While
the results section does reference the outcomes of some OLS regression models, the full results are
never shared. Further, my understanding of the analyses performed suggests that OLS is not the
appropriate regression model for the data being analyzed. My understanding is that the dependent
variable in all regression models is a count variable. In such instances the regression models should
employ either the Poisson or negative binomial models, depending upon whether the data are
overdispersed (Cameron & Trivedi, 1986). Such models recognize that the distribution of count
dependent variables cannot take on negative values. As such, I have little faith in the results reported
using OLS and the analyses should be re-done using on the above models (with tables presenting the
results, of course).
RESPONSE: Thanks for raising this important point. Since we included only one control variable in our
models, we decided to include the results only in the text. We felt that this approach helped to conserve
valuable manuscript space. For your references, though, we are including the regression results in an
appendix to these responses.
You are correct that our dependent variable is a count variable. As such, we should rely on Poisson or
negative binomial models. In supplementary analyses that we now include in a footnote, both
approaches yield virtually the same results. We decided to continue to rely on OLS regression in the
body of the paper, however, because the coefficients make more sense intuitively (which we discuss). In
contrast, Poisson or negative binomial results are expressed as likelihood values. We hope you find our
approach acceptable. Thank you for the suggestion, because it helps a great deal to illustrate that our
results remain consistent across a number of analytic techniques.
5. COMMENT: As it relates to the actual implications of this study the authors wrote, “the pool of
suitable candidates for the micro position is nearly three times larger than the pool for similar macro
candidates” (17). This makes the assumption that tenure requirements for the number of top journal
hits is identical for micro and macro researchers and, in my experience, this is simply not true. Based
upon some phone calls I made to research universities, macro researchers can get tenure with 4 or even
3 hits in top journals where the expectations are higher for micro researchers. I think this is a point the
author(s) should mention and discuss. Running some sensitivity analyses may help on this point. What
happens when the micro researchers must publish 6 articles and macro researchers must publish 4? If
the results still hold, I think the author(s) have a stronger case to make that the size of suitable
candidates for each discipline are significantly different.
RESPONSE: We now elaborate on this point more in our discussion section. We are reluctant, however,
to take a strong position on this point. In our experience working at research universities, there is not a
sliding scale in terms of tenure requirements based on research focus. In fact, we also are not aware,
based on our experience, with the publication rates that you specify for macro scholars being sufficient
Furthermore, we are hesitant to change the assumptions of our process for the following reason. One of
the objectives of our paper is to compare the relative productivity of macro and micro scholars. This
becomes more difficult if one group is held to one standard while another group is held to a different
standard. Moreover, we do not want to appear as if we are arguing for less rigorous tenure
requirements for one group as compared to the other. Instead, we are reporting the data and hope the
readers will reach their own conclusions.
6. COMMENT: Further, the field has recognized the shortage of macro journals with the addition of
Organization Science in 1990 (which, based upon my phone calls, is universally considered a top journal)
and Strategic Organization (which is not yet considered a top journal, but may be in the future). While I
realize the time period of the study did not consider top journals, to the extent that conclusions are
drawn about the current state of academia, such changes must be recognized and discussed. Indeed,
the founding of such journals seems to be a response to the findings of the current study.
RESPONSE: Per your suggestion, we have included Organization Science as a journal in our sample. We
also now discuss the notion of new journals in our discussion section. Thank you for the great
7. COMMENT: Lastly, it seems that this article may not be a great fit for AMLE. Whenever I read an
article and this suspicion hits me I always check the references to check the rate at which the target
journal is cited. In this case only 3 of the roughly 30 references are from AMLE. This confirmed my
suspicion and suggests that AMLE may not be the proper target for this article in its current form.
RESPONSE: Thank you for this comment. Thankfully the editor provided us with a great deal of guidance
regarding how to better position the article to suit the mission of AMLE. Part of this guidance involved
pointing us to related articles that have appeared in AMLE. We are hopeful that you agree that this
revision is better targeted toward AMLE.
8. COMMENT: Overall, I think there is a potential contribution here, but the manuscript must be revised
substantially in order to do so. In its present form it does not seem to fit AMLE very well, I am not
confident the statistical analyses were performed properly, and I am dubious of some of the conclusions
drawn I the discussion given the current state of the field.
RESPONSE: Thank you once again for your valuable comments and suggestions. We are hopeful that our
supplementary analyses and revised focus helps to improve your opinion of work.
Appendix: Full Regression Results
Intercept -21356*** -22.581*** -22.456***
(2566.96) (3.196) (2.760)
Year 10.70*** .014*** .014***
(1.31) (.002) (.001)
Munificence 1.95*** .002*** .002***
(.19) (.000) (.000)
Documents you may be interested
Documents you may be interested