Rankings of Academic Journals and Institutions
in Economics∗
Pantelis Kalaitzidakis
University of Crete and University of Cyprus
Theofanis P. Mamuneas
University of Leicester and University of Cyprus
Thanasis Stengos
University of Guelph
October 2001
∗We like to thank D. Laband who provided us with his index of characters per page and
Elena Ketteni for excellent research assistance. Financial support from European Economic
Association is gratefully acknowledged.
1 Introduction
There has been a lot of recent research literature on rankings of economics
departments throughout the world. They serve as signals tools for attracting
new faculty and retaining older in highly ranked institutions and also help at-
tract the best graduate students who have academic aspirations. Many times
these rankings are used by university administrators to allocate scarce educa-
tion funds to different departments according to their success in these rankings.
There has been a long standing tradition for US economic departments to be
ranked (see Scott and Mitias (1996) and Dusansky and Vernon (1998) for recent
such rankings). Recent European studies of this kind include Kirman and Dahl
(1994) and Kalaitzidakis, Mamuneas and Stengos (1999). There have been also
rankings of departments in Asia (see Jin and Yau (1999)), Canada (see Lucas
(1995)), as well as Australia (see Harris (1990)). Rankings are also constructed
in other related disciplines such as finance for the same reasons outlined above,
(see Chung and Cox, (1990)).
Coupé (2000) provides a comprehensive ranking of economic departments
world-wide. His ranking methodology is based on employing various perfor-
mance measures from the existing literature, such as the citations weighted
journal ranking by Laband and Piette (1994), to assess the output of individ-
ual researchers and then according to their affiliation compute the department
rankings. He reports the rankings from the different methodologies and he also
presents a ranking based on the average of these different methods. However,
the latter ranking is based on averaging rank statistics and as such it is not very
informative.
A common drawback that permeates most of the studies that produce de-
partment rankings is that they are based on a certain ranking of economic
journals that was itself constructed over a certain time period that typically is
different from the corresponding period of the department rankings. Hence, a
typical list of journals that is citations weighted uses weights that correspond
to an earlier period from the current one. That means that the most current
research outlets that are used by the profession (new journals, improved older
journals etc.) are not used with their true weights for the period under inves-
tigation. Hence, potentially rankings that use a list of research journals with
weights from a different period may produce biased and unreliable rankings for
the current period. In this paper we try to rectify this deficiency in the litera-
ture by both computing an updated list of journal rankings with current weights
computed from their citations impact and then use those to produce a world
1
wide ranking of academic institutions.
The paper is organized as follows. The next section provides the methodol-
ogy that we employ to arrive at the new journal rankings. We provide details
of the way that we arrive at these journal rankings that form the weights to be
used for the derivation of the institutional rankings as well as the methodology
that is used to construct the latter. In the next section we discuss the results.
Finally we conclude.
2 Methodology
2.1 Journal Rankings
The ranking of economics departments based on research output requires two
important ingredients. First, the choice of the set of research output outlets,
journals and second the choice of the weights to adjust the different journals in
terms of quality, age and size.
In our proposal the set of journals we choose consists of the thirty top eco-
nomic academic journals based on the number of 1998 citations of articles pub-
lished in previous periods. There are already some relatively recent rankings of
journals based on 1990 citations of articles published in 1985-1989 by Laband
and Piette (1994). However, we felt that these rankings should be updated given
the rapid expansion of publications, new entrants, and changes in emphasis in
the profession. In fact our findings suggest that the earlier journal rankings do
not accurately reflect the current trends in the profession and hence all existing
studies using them as a basis of constructing department rankings would lead to
unreliable and inaccurate results. Below we outline in more detail the methodol-
ogy we have employed in arriving at a more representative and accurate journal
ranking.
One source of valuable information of the citations received by the economic
journals is Journal of Citation Reports (JCR). JCR also ranks economic journals
based on the number of citations received. For instance in Table 1 last column
we report the ranking of economic journals based on the number of citations
received in 1998 of articles published in previous years (more than 10 years).
We have standardized the top journal, American Economic Review to be equal
to 100. This ranking is based on the category “economics”. Note also that
we have exclude journals that are not academic, for example The Economist.
That does not include journals that are core journals in other related disciplines,
such as the Journal of Finance. However, the Journal of Financial Economics
2
is included in this category.
Even though this ranking as a first approximation seems reasonable, is in
general unsatisfactory for the following reasons: a) Self-citations are included,
something that biases the rankings (due to the common tendency of journals
to cite their own articles more often). b) There is no correction for the age
of a journal (older journals tend to accumulate more citations). c) “Bigger”
journals that tend to publish more articles, also attract more citations, and
most importantly d) citations are not adjusted for the impact that the most
influential journals have on the profession.
In order to correct for self-citations and the age of a journal we have con-
structed a new ranking of journals by excluding self-citations and all the cita-
tions of articles published before 1994. In other words, the index of the seventh
column is based on citations of 1998 of articles published in 1994-1998 period
excluding self-citations (see 6th column of Table 1).
The rest of our ranking of journals is based on citations of 1998 of articles
published in 1994-1998 period excluding self-citations and adjusted for the im-
pact (influence) and size. To correct of the impact of the journal we have broadly
followed the methodology of Liebowitz and Palmer (1984) (see also Laband and
Piette, 1994). This methodology is based on an iterative procedure which we
briefly outline below.
Let Cij be the number of citations to journal i from journal j, n the number
of journals in our list, Zi a factor adjusting for the size of a journal and δj taking
values one usually or zero when there is no information and it will be discussed
shortly. The t iteration is given by
Ii,t =
Pn
j=1 δjCij
Zi
Ij,t−1,
where
Ii,0 =
Pn
j=1 δjCij
Zi
This process usually converges after 10 to 15 iterations The results reported in
Table 1 are based on 50 iterations. The adjusted rankings for self-citations and
age of journal are presented in column 4 of Table 1.
Column 5 of Table 1 reports the ranking of journals by the impact adjusted
citations without adjusting for the journal size, i.e., Zi = 1 (δj = 1 for all
journals). The rest of the columns are based on different types of Zi. These
are the average number of articles a journal had published in the period from
1996-1998 (this was the only available information of JCR), the average number
of pages published in the same period taken from ECONLIT, and finally the
3
number of characters published. The total number of characters published per
year is calculated by the number of characters per page times the average number
of pages published. An index of character per page (American Economic Review
equal to one) for seventy journals were made available to us by Laband and
Piette and has also been cross-checked and supplemented with our calculations.
In total we are able to have information of the characters per page for ninety
two journals. For the journals that we do not have information we set δj = 0
(otherwise δj = 1). Thus we do not count these journals as a source of citations
but we count them as receivers. Note that this does not constitute a large source
of bias for our rankings of the top journals since the lack of information about
the characters per page is concentrated in the lower ranked journals, where the
impact contribution is very small. (It can also be verified that rankings do not
change much using the different measures of size).
It is interesting to note that comparing the rankings of Table 1 with the
previous journal rankings of Laband and Piette (1994), the relative positions
of top journals have not changed much. However, the weights have changed
considerably. It seems that the distance of most journals from the American
Economic Review has increased, with the notable exception of Econometrica
which now appears to be the leading economic journal when we use characters
per page as a measure of size. In addition, more empirically oriented or applied
journals have risen in the rankings, e.g. the Journal of Business and Economic
Statistics and the Journal of Applied Econometrics. A surprising result is the
appearance of Econometric Theory and Economic Theory in the group of the
30 top journals, when we correct for journal impact. One possible explanation
is that these journals receive a lot of citations from top ranked journals like
Econometrica, Journal of Econometrics and Journal of Economic Theory re-
spectively. Finally, it is interesting to note that the European Economic Review
has risen considerably in statue and it is included in the group of thirty top
journals, while in the study of Laband and Piette (1994) it occupied the 50th
position. Note that there is an overall agreement between all rankings methods
for the top group of journals at least as far as the composition of this group is
concerned.
2.2 Institutional Rankings
The analysis is based on publications in the top 30 journals according to our
pages adjusted rankings (column 2 of Table 1) for the five year period 1995 to
1999. Given that we lacked information on characters per page for the whole
4
set of journals and given the similarities of the weights of both journal rank-
ings methods (columns 1 and 2 of Table 1), we opted for the construction of
institutional rankings using pages as the measure of size. The selection of the
top 30 journals provides a rich group of research outlets for the core of eco-
nomic theory and econometrics as well as the most respected field journals. It
is an updated ”Diamond List”, (see Burton and Phimister (1995)) that has
been extensively used in the rankings literature as the standard list of quality
journals. The last journal that is included in the list, the Journal of Interna-
tional Economics, has an impact factor of 0.0784 compared with 1.00 for the
American Economic Review. The list of journals that are included account
for more than 90% of all citations. For these journals there is a broad agree-
ment among all ranking criteria that they belong to the top group, see Table
1. The impact factors for the journals that are excluded from the list are quite
small and even if they were included in the calculations they would not make
much difference in the overall construction of rankings especially for the top
200 economic departments that we report. The included journals are: Ameri-
can Economic Review, Journal of Political Economy, Econometrica, Quarterly
Journal of Economics, Journal of Econometrics, Journal of Economic Perspec-
tives, Journal of Economic Theory, Journal of Monetary Economics, Review of
Economic Studies, Review of Economics and Statistics, The Economic Journal,
European Economic Review, Games and Economic Behavior, Journal of Busi-
ness and Economic Statistics, Journal of Public Economics, Journal of Human
Resources, Journal of Economic Literature, Econometric Theory, Journal of
Labor Economics, International Economic Review, Economic Theory, Journal
of Environmental Economics and Management, Rand Journal of Economics,
Journal of Financial Economics, Economics Letters, Journal of Applied Econo-
metrics, Oxford Bulletin of Economics and Statistics, Scandinavian Journal of
Economics, Journal of Economic Dynamics and Control and the Journal of
International Economics.
We allocate article pages to the affiliation of the authors at the time of publi-
cation. Affiliations taken from the published articles reflect the actual research
output produced, in contrast to the current affiliation of the authors which
might serve as a proxy for future research output for the institution where the
researcher currently resides. In papers with n co-authors, each co-author is al-
located 1/n pages of the article. In addition, when m affiliations are listed by
some author, then we allocate to each affiliation 1/m of the pages that corre-
spond to the specific author. We do not include among the various affiliations
those that correspond to certain research centers that act as umbrellas for var-
5
ious researches but do not offer a permanent home base, such as NBER in the
USA and CEPR in the UK. In case authors include the above as joint affiliations
then all the weight is attached to their primary affiliations. We also excluded
from the calculation of rankings the research output that is produced primarily
at non academic centers such as the various central banks, the World Bank and
the IMF. Since our primary task is to evaluate research carried out at acad-
emic institutions including non academic research centers would not constitute
a valid comparison, since academics usually have also teaching duties that oc-
cupy much of their time. We have included as part of the institutional research
output the published research that has been produced by faculty members of
business schools that belong to these institutions. That gives an advantage to
institutions with large vibrant business schools, such as the top US universities.
However, since our task was to record the research output in economics carried
out in academic institutions in general, excluding business school output would
have left out a significant part of current research. For the same reason we also
include as part of a given institution research centers that are located in these
institutions and are frequented by researchers. For example, the Institute for
Fiscal Studies (IFS) has been included as part of University College, London.
There has been a trend in the recent literature, see Baltagi (1999), Coupé
(2000) to also produce rankings of individuals in the same way as institutional
rankings are produced. In so far as these individual rankings simply state the
number of total pages published by individuals we are not sure that they ad-
dress the issue of impact in the profession that various individual researchers
may have. Institutional rankings are based on citations adjusted pages with the
adjustments factors coming from overall averages. Yet an individual researcher
has an impact on the profession because of his specific contribution. To conduct
a proper and meaningful comparison of individuals one should look at the cita-
tions of specific articles that each researches has published. Also in that case
one would like to take a long-run view of this impact and hence examine the
rate of citations over time, something that is not apparent from a total number
of published pages calculation. To offer such a ranking would require tracking
down each individual’s citations record, something that is well beyond the scope
of the present study.
6
3 The Results
3.1 World Rankings
Table 2 presents the world-wide rankings of economic departments. Since the
current literature is quite exhaustive in the construction of rankings with ad-
justments based on previous studies (see Kalaitzidakis et al (1999) and Coupé
(2000)) we only present the rankings based on the current impact factors for the
list of the 30 journals that we discussed in the previous section. The first column
presents rankings based on the adjusted pages produced using as weights those
of column 2 of Table 1. Concentrating on a single methodology gives a clearer
impression about the institutional standings. Presenting results with different
methodologies and then averaging out the different ranks obscures the trends
that are taking place in the research output of the profession. Column 3 of Table
2 simply presents the unadjusted total pages produced by each institution and
column 4 the weighted adjusted pages using the weights from the 2 column of
Table 1 for the chosen set of journals. Table 2 presents the top 200 institutions
in the world.
There are some very interesting facts that emerge from Table 2. The US
institutions are not in the majority (they constitute 44 percent of the total
(87 placements in the group of 200)). The European affiliations constitute 35
percent. Including Israel among the European institutions as in Kalaitzidakis et
al (1999) raises the above number to 38 percent. There is 8 percent allocated to
Canadian institutions (15 institutions). The Asian profession shows a credible
presence with 8 percent or 15 institutions in the top 200 group. The rest is
made up from 6 universities from Australia, 1 from New Zealand and 1 from
Chile. The picture is more skewed towards the US dominance if one looks at the
group of the top 50 universities, where the US schools make up 70 percent of
the total. In that group there are 7 European institutions (9 if one adds the tow
universities from Israel), 5 Canadian and 1 from Hong Kong. Harvard, Chicago
and MIT make up the top three universities. There are 18 US schools in the
top 20 with only Tilburg University and the London School of Economics make
it in the group of the top 20.1 The US presence falls to 54 percent in the group
of the top 100 and further to 44 in the top 200. In that case the European
presence (with the inclusion of Israeli schools) doubles from 16 to 33 percent
for the top 100 group and it increases further to 37 for the top 200 group. It
1 It should be noted that the top US institutions benefit from the presence of very strong
business schools. A lot of economic research takes place at these business schools. In Europe
business schools typically stand on their own as separate entities.
7
seems that it is in the group between 50 and 100 that European universities have
improved and are doing relatively quite well. It is the group that consists of
many institutions that are producing excellent research that competes favorably
with their US counterparts who belong to that group. In previous studies that
only considered the top 20 North American universities, it was asserted that
Europe was lagging behind significantly North America and the US in particular
in terms of research produced by the top regional institutions respectively, (see
Kalaitzidakis et al, (1999)). This may be true for the top 20 institutions as
it was noted earlier but as a general statement this is simply not true as it is
evident from the discussion above. In the comparison that takes place after
the group of the top 20, European institutions are overall at par in terms of
research output with their US counterparts. Furthermore, it is interesting to
note the presence of the Asian universities that appears in the group of the top
200. In particular, we note that one university f