Printer Friendly

Production in the finance literature, institutional reputation, and labor mobility in academia: a global perspective.

Academic institutions are ranked on a global scale in terms of finance literature productivity. US institutions are dominant in academic publishing although European and Asian institutions have improved sign significantly in recent years. Additionally, we study the relationship between the quality of human capital and the likelihood of an upward career move. Our results show that an individual relocates to a higher-ranked institution exhibits a research record that is two times stronger than that of an average faculty member at the destination institution. We further model the probability of an upward move in the academic labor market as a function of human capital using an ordered logistic model. We find that publications in sixteen core journals, publications in three top journals, and the rank of the Ph.D. granting institution enhance the probability of moving to a higher ranked institution. On the other hand, the length of teaching experience decreases this probability.

**********

Research productivity has always been a topic of substantial interest to the academic community. These studies can be found in many disciplines. Some examples are Niemi (1987) and Borokhovich, Bricker, Brunarski, and Simkins (1995) in finance, Brown (1996) in accounting, Conboy, Dusansky, Drukker, and Kildegaard (1995), Scott and Mitias (1996), and Collins, Cox, and Stango (2000) in economics. There are several possible reasons for the interest in this topic. First, promotion/tenure, faculty compensation, and resource allocation decisions in academic institutions are often dependent on the productivity of the faculty, and the quality and quantity of publications that the faculty produce are factors used to measure this productivity (Gomez-Mejia and Balkin 1992). Second, the ranking of academic institutions draws attention from faculty position applicants, college and graduate school bound students, and potential university donors. Popular publications, such as US News and World Report, offer guidance to thes e individuals by ranking colleges and universities with academic reputation among the ranking criteria. A recent article in the Chronicle of Higher Education (June 1, 2001, p.A8) reports that Ohio State University "taxes" departments in order to distinguish selected departments within their school as "top-notch". "Administrators (at Ohio State) believe the rankings are important because they give graduate students, the public, and state legislators a benchmark. 'They are an indication of what league you're in,' says Randall B. Ripley, Dean of the College of Social and Behavior Sciences." (p. A9)

Despite the interest in ranking academic programs, the assessments have consistently focused upon North American institutions. Moreover, few studies provide insights into other related issues beyond the ranking outcome. Therefore, the objective of our research is twofold. Our first objective is to study research productivity within the academic discipline of finance using a global sample of institutions. Globalization has reduced barriers in almost every sector of the economy and the academic arena is no exception. Moreover, an examination of ranking academic departments on a global scale is not without precedence. For example, in January 2001, the Financial Times published its world rankings of the top 100 MBA programs. However, their ranking study is confined to MBA programs, with data gathered through opinion surveys from participating MBA programs; consequently, their study does not provide an assessment of finance programs on a global scale which is one contribution of our study.

Our second objective is to study labor mobility within the finance profession. To the best of our knowledge, this has yet to be investigated. Although there are many factors that affect an individual's ability or desire to relocate and to receive an offer, our focus is limited to a more narrow aspect of labor mobility. We focus on how an individual's human capital impacts that individual's occupational mobility. The quality of human capital is difficult to quantify for many occupations, but in academia it is more feasible. This is because one of the major quality indicators in academia is a researcher's publication record and that record can be objectively quantified and measured. Consequently, our study allows us to examine the role that a researcher's publication record plays in moving from a lower ranked institution to a higher ranked institution. We use alternative research publication measures to proxy the quality of an individual's human capital. These measures include, for example, the total published pages in a set of 16 core finance journals, the per year published pages from the same set of finance journals, and the weighted number of articles in the top-3 finance journals.

Besides the uniqueness of our study in developing global rankings and in investigating labor mobility, further strengths of our study include a substantial sample period and a careful accounting of articles with respect to co-authorships, co-affiliations, and differences in average article lengths across journals. Our sample period extends from 1990 through 2001 while prior studies that rank institutions in finance have typically used more limited sample periods (e.g., 4-5 years). A benefit of our longer sample period is our ability to compare the research productivity during sub-sampling periods and, therefore, to assess a university's improvements in research productivity over time.

There are, however, caveats in our database and the derived results. First, because it is possible that a small number of authors from other disciplines may have contributed to the finance literature, the rankings of certain finance departments may be overstated. Second, because some individuals with a strong background in economics also publish finance research in elite economics journals, the lack of inclusion in our sample of these types of publications is likely to understate the total research productivity of some institutions.

The results of our research offer the following conclusions: 1) Top-ranked finance programs are still dominated by North American institutions. In effect, only three non-North American institutions (London Business School, Hong Kong University of Science and Technology, and City University Business School) are in the top-SO ranked institutions. 2) Non-North American institutions are catching up. There are a number of foreign institutions showing substantial increases in absolute total pages published over the second half of the sampling period. 3) There is a positive relationship between research variables that proxy the strength of an individual's human capital and the ability to move to a higher ranked institution.

The remainder of this research is organized as follows: Section I describes our database. Section II presents our findings on the global rankings of institutions and the changes in research productivity over time. Section III discusses and empirically analyzes labor mobility as a function of variables that proxy for the strength of an individual's human capital. Conclusions are offered in Section IV.

I. Sources of Data and Descriptive Statistics

The global research productivity data is hand collected using the hard copies from a set of 16 core finance journals from 1990-2001. (1) The data set includes each author's name, his or her affiliation, and the total page count of each article. These particular journals were chosen for several reasons. First, an almost identical set of journals has been used in prior studies (e.g., Borokhovich et al. 1995). (2) Among these 16 journals, many of them have been documented as the most influential journals in finance (see Alexander and Mabry, 1994). Second, in order for us to measure the longer-term effort by the institutions, the research outlets (the journals) need to exist for a minimum of 12 years. This screening process excludes some highly regarded newer finance journals (e.g., Journal of Empirical Finance). Third, the 16 journals include specialized journals, such as Journal of Futures Markets, Journal of Portfolio Management, and Journal of International Money and Finance, as well as journals with a general scope. Hence, we are able to measure the research productivity of faculty with respect to their different research interests.

While our database spans a 12-year period that allows us to conduct a more comprehensive analysis, there are caveats in the database and the derived results. First, while faculty in finance departments are the authors for most of the articles published in this set of journals, it is possible to find a small number of contributions made by authors from other disciplines, particularly economics, statistics, and/or accounting. This may be the case in selected situations, because some universities do not have finance departments. Moreover, on many occasions, the authors' department affiliations are not specified in these journal articles and there are no reliable sources to identify the exact departmental affiliations of all authors. Therefore, the ranking performed in this study may overstate the ranking of a particular finance department. Nevertheless, we believe that publications of finance articles by other disciplines in the same university also contribute to the general reputation of a finance program in a university. A second related concern is that some authors also published finance research in economics, accounting, statistics, and other journals. It is practically impossible for us to search every article in a very large set of journals in other disciplines and decide which articles represent contributions to the financial literature. Exclusion of these papers undoubtedly will understate the true human capital of these researchers. For example, given the overlapping research topics in economics and finance, some individuals with strong background and interest in economics often also publish finance research in elite economic journals such as American Economic Review and the Journal of Political Economy. Since these people are disproportionately based at top-ranked institutions, the total publication record of these institutions may be understated.

Similar to earlier studies, three measurement adjustments have been made to the raw data. First, in cases where there are more than one author, we calculate a weighted number of articles published per author by dividing the article by the number of authors. Moreover, while it is not clear whether previous studies consider possible co-affiliation of an author, we make adjustments in such cases. Specifically, when an author is affiliated with more than one institution, the contribution is divided equally among the stated institutions.

Second, manually collected data from each journal are checked for accuracy. It is quite common that an author uses slightly different names over his/her career. In cases where author names are similar, but not identical, we conduct additional research to verify the identity of the authors and standardize the presentation of his/her name. In a few cases, there are name changes for the same university. We convert all the "different names" into a single name in such cases.

Third, we convert total page counts in each journal to JF-equivalent pages. Specifically, we randomly select three full text pages (no equations, no footnotes, and no graphs) from each journal and count the number of words on each page. Based on the counting and the three-page average, we obtain the number of words on a "typical" page for each journal. Next, we calculate an adjustment factor by using the words on each "typical" page in each journal as compared with a "typical" page in the Journal of Finance. For instance, the Journal of Finance has 582 words on a typical page while a typical page in the Journal of Business has only 492 words. Therefore, the Journal of Business total page count is converted to JF-equivalent pages by multiplying the Journal of Business page counts by 0.84536 (i.e., 492/582). (3)

Although there are different ways to rank the research productivity of an institution, we chose the JF-equivalent page count method for the following reasons: 1) A popular alternative quality measurement, the Social Science Citation Index (SSCI) impact factor, is not available for all finance journals. (For example, the index is not available for the Financial Analysts Journal.) 2) A finance journal's SSCI impact factor, even if available, may be affected by citations outside the finance field (Borokhovich, Bricker, and Simkins, 2000). 3) Citation factors often are biased in favor of subject matter in the corporate finance area (Borokhovich et al., 2000). 4) Although adjusted article counts are also reported, we believe that adjusted page counts serve as a better measurement. Borokhovich et al. (2000) find that, in general, shorter papers have less influence than full articles. Thus, we can mitigate the issue of differential quality across journals by using JF-equivalent page counts. Because papers published in the top tier journals are, on average, longer than articles in other journals, the adjusted page counts method implicitly takes journal influence into account. For example, a typical full-length Journal of Finance article runs 35-40 pages, while a typical Financial Review article has approximately 15 pages (or 11.5 JF-equivalent pages). 5) We use an institution's total adjusted page counts rather than "per capita productivity" because we believe institutional reputation is measured by its aggregate academic power to which total productivity contributes more strongly than per capita productivity.

In total, during the period of 1990-2001, the l6-journals contain 8,029 articles written by 6,538 authors from 923 universities and 860 non-academic institutions with 143,543.30 JF-equivalent pages. For the academic affiliations (including co-affiliated authors), there are 4,990 authors who wrote 6,536.35 weighted articles with 121,055.42 JF-equivalent pages.

Table I presents the summary statistics of research productivity by academic institutions and by academic authors. In this Table, we report the weighted number of articles published, the unweighted number of articles published, and the JF-equivalent page counts published as the productivity indicators. In Panel A, the mean values of the weighted number of articles, the unweighted number of articles, and the JF-equivalent pages per academic institution are 7.08, 14.38, and 130.52, respectively. Judging from the significantly smaller median values, the distribution is highly skewed. For example, the median JF-equivalent page per institution is 25.11 pages, which is much smaller than the mean of 130.52 pages. The skewness and kurtosis statistics are all positive and large for all three research productivity measures.

Panel B of Table I summarizes research productivity by authors affiliated with academic institutions. Some of these authors may have both academic and non-academic dual-affiliations. From Panel B, an average author would produce 1.31 weighted articles, 2.53 unweighted articles, and 24.26 JF-equivalent pages during the period of 1990-2001. Similar to Panel A, all three research productivity measures also suggest the positively skewed and very peaked empirical distributions.

Panel C gives further insight into the frequency of publications for individual authors. Among the 4,990 academic authors, 2,719 (54.49% of total) have published only one article (unweighted) during the 12-year period. When we add to that percentage those who have published two unweighted articles (17.15%), we find that more than 70% of all authors have published two articles or less during the 12-year period. Publishing seven or more articles in twelve years, therefore, places a researcher in the top six percentile of the productivity distribution. This result is similar to the findings of Zivney and Bertin (1992) who conclude that for any individual to publish one finance article per year in a larger set of academic journals would place him/her in the top five percent of all authors. (4)

II. Global Institutional Ranking

We rank academic institutions based upon research productivity measured by publications in 16 finance journals as well as in a sample of the "top-3" journals. We also examine the progress by academic institutions of the research productivity over the 12-year period.

A. Global Institutional Ranking Based upon All 16 Journals

In this section, we rank academic institutions on a global scale according to their research productivity. The rankings are presented in Column 1 of Table II. (5) The rankings are based upon JF-equivalent page counts. (6) Table II provides the top 100 institutions and the related information. The results of the remaining 823 institutions, not reported in this paper, are available on the authors' website. Columns 1 and 2 provide overall ranking of the institutions with country names (other than US) inside the parentheses. Not surprisingly, the North American and Canada) universities dominate the top 100 ranking. Indeed, no universities outside of the US are in the top 20, and the highest rank for a Canadian university is the University of British Columbia at 24th. Moreover, the highest-ranking European university is London Business School that places 22nd and the highest ranked Pacific-Basin university is Hong Kong University of Science and Technology that places 40th. Among the top 100 universities, only 15 universities are outside North America (9 in Europe, 2 in the Middle East, and 4 in the Pacific-Basin). We also provide a diagram to show the distribution of the JF-equivalent pages of all 923 universities that published at least one article in the 16 finance journals. Figure 1 shows a plot of the cumulative percentage of JF-equivalent pages against the cumulative number of universities (arranged from highest to lowest rank). The cumulative distribution is highly skewed as evidenced by the initial steep slope and the subsequent flattening of the curve. Specifically, the top 11 universities account for 20% of all the JF-equivalent pages. The top 37, top 84, and top 179 universities account for approximately 40%, 60%, and 80% of all the JF-equivalent pages, respectively. This highly skewed cumulative distribution of research productivity is in line with recent findings in citation studies that suggest a small number of articles dominate the total citation counts (see Chung, Cox, and Mitchell, 2001).

We also examine the progress of the research productivity of the same 923 universities over the twelve years period. To examine this progress, we compare the research productivity of the first six years (1990-95) with that of the last six years (1996-2001) of the sample. The results are in Columns 5 and 7 of Table II. Rankings within each sub-period are reported in Columns 4 and 6. Column 8 presents the absolute changes in JF-equivalent pages over the two sub-periods and Column 9 contains the ranking based on the improvement in the research productivity. (7) Based on Column 9, the top five most improved universities are Hong Kong University of Science and Technology, University of Pennsylvania, Yale University, London Business School, and Georgetown University. Although research in finance is still dominated by the North American institutions, there is a clear trend in the growing influence of European and Asian universities that may be expected to play an even larger role in the future. (8)

B. Global Institutional Ranking Based Upon the Top-3 Journals

While Section A reports research productivity measured by publications in a set of 16 core journals, many major research universities give heavier weight to publications in the top three journals. Therefore, we also rank institutions based only upon the research output in Journal of Finance, Journal of Financial Economics, and Review of Financial Studies. These results are reported in Table III, Columns 1 through 3. During the 12-year sample period, 319 institutions have published at least one unweighted article in these three journals, and the remaining 604 institutions did not. It should be noted, however, that the institutional JF-equivalent page counts based upon the 16 journals are very highly correlated with the page counts based only upon the top-3 journals for these 319 institutions. Indeed, the rank correlation coefficient is 81.26%. Therefore, the top-ranked institutions do not differ significantly when the analysis is based upon all 16 journals or based only upon the top-3 journals if the institut ions published at least one articles (unweighted) in the top-3 finance journals. When comparing Tables II and III, the top ten universities are the same with a slight difference in relative ranking. Specifically, New York University and the University of Pennsylvania trade their 1-2 positions and Duke University and Ohio State University swap their 9-10 positions. Similar to the results reported in Table II, only 14 universities ranked in the top-100 are outside North America (9 in Europe, 3 in the Pacific-Basin, and 2 in the Middle East).9

Similar to Table II, we also re-examine the ranking of academic institutions based on the top-3 finance journals in the 1990-95 verse the 1996-2001 sub-periods. The results are presented in Table III Columns 4 to (9). Based on Column 9, the top-5 most improved universities are New York University, the University of Pennsylvania, Stanford University, the University of California-Davis, and the University of Maryland.

III. Human Capital Endowment and Labor Mobility

In this section, we study the relationship between variables that proxy the quality of human capital and labor mobility in the finance discipline. Based on our database from 1990 through 2001, we are able to identify faculty who made career moves over the period by examining whether affiliation changes exist. For the purpose of our study, the institutions are classified into five different ranking levels according to their research record measured by their JF-equivalent page counts as reported in Table II. We define Level 1 institutions as institutions that belong to the top 20 percentile of total JF-equivalent page counts. Because the JF-equivalent page count distribution is highly skewed, eleven institutions account for 20% of the total publications. The next 26 institutions published an additional 20% of the total JF-equivalent page counts and, therefore, are defined as the Level 2 universities. Following the same logic, institutions ranked from 38 to 84 are the Level 3 universities, institutions ranked fr om 85 to 179 are the Level 4 universities, and the remaining 744 institutions are the Level 5 universities.

Essentially, there are three types of career moves with respect to the ranking classes of the departing school and destination school: lateral move, downward move, or upward move. However, there is one caveat. We cannot identify individuals who may have moved but stopped publishing in the 16 core finance journals after their job changes. Panel A of Table IV reports preliminary statistics. Among the 5,268 identifiable individual author-affiliations in academia, (10) 1,549 did not change affiliations during the twelve years period; 254 made a lateral movement; 384 moved to a lower-level institution; 362 moved to a higher-level institution; 2,719 published only one article during the twelve years period and, therefore, we could not identify if a change in affiliation did occur. Because our analysis focuses on those individuals who moved to a higher-level institution, we do not believe individuals who have published only one article in twelve years have the human capital requirement to move to a higher-level ins titution. These 2,719 unidentifiable individuals, therefore, are not likely to be the upward movers and thus should not bias our analysis. Panel A of Table IV also reports the average number of publications across these categories. For the "non-movers", the average number of unweighted publications is 3.76. For the "movers", the average number of unweighted publications ranges from 5.67 to 5.91 articles. Similarly, the average number of weighted articles and JF-equivalent pages for the "non-movers" are also less than the "movers". Moreover, at the first glance, the results from Panel A show that more human capital is needed in order to move downward than upward since the average number of weighted articles, unweighted articles, and JF-equivalent pages are larger for the downward movers. This puzzle may be explainable. For example, an individual moving down from a Level 1 institution to a Level 3 institution may have more human capital than an individual moving up from a Level 4 institution to a Level 3 instit ution. Nevertheless, the results from Panel A are based upon an unfiltered sample, and we explain in the following sections how we further screen our sample for a cleaner definition of affiliation changes.

In Panel B, we further report the publication record for those researchers who have not changed affiliations during the 12-year period, classified according to the levels of their employer. It is not surprising that all measures of research productivity increase monotonically for the individual researcher, as the rank of the institutions increase. More interesting, however, will be the comparison of these statistics with the same statistics for those researchers who relocated to higher ranked institutions as reported in Table V.

Before reviewing Table V, however, we present the issues that necessitate the filtering of the data set. First, since not all affiliation changes are career moves, we need to exclude those affiliation changes that are visiting appointments, part-time appointments, and moves between non-finance disciplines. These types of job changes may be expected to have different human capital endowment requirements. (11) Second, after this screening process, our study is further confined to finance faculty who permanently move from a lower-level institution to a higher-level institution in a tenure or tenure-track position within the finance discipline. We focus only on upward movers since we cannot unambiguously relate human capital to downward (and lateral) movements because higher levels of human capital may not be a necessary condition for a downward movement as there are non-research reasons contributing to downward (and lateral) career moves. A faculty member with substantial human capital may make a downward or a lateral career move to take a more prestigious position or to live in a specific geographic location. (12) On the other hand, downward movers may change affiliations to a lower ranked institution due to a lack of required human capital. Yet, it is impossible to distinguish these two groups of downward movers ("non-research related" movers or "lack of human capital" movers) without controlling for other factors that contribute to the move. (13) For a faculty member to move to a higher-ranked institution, however, human capital is a necessary condition for the move irrespective of other non-human capital related motivations. Consequently, our sample is restricted based upon the following selection screening process: 1) we only include authors that change to a higher-ranked institution; 2) we only include authors that have verifiable tenure or tenure-track appointments in the post-movement institutions; and 3) we only include authors from whom we can verify the Ph.D. granting institutions, the exact year of the relocation, and the length of teaching experience.

This screening process yields 88 qualifying cases. In order to identify the relevant information set for this group of researchers, we examine personal web-pages of the authors, James R. Hasselback's (1999) Finance Faculty Directory, Heck's (1998) Finance Literature Index, the World-Wide finance faculty directory at Ohio State University, and the membership directory of the Financial Management Association International. In certain instances, we also sent e-mails to faculty for whom we could not verify whether they actually changed affiliation and/or the year they changed affiliation.

In Table V, we provide summary information on the sample of 88 cases of upward career moves that met our selection requirements. Among these cases, there are 16 faculty that moved to Level 1 universities, 17 to Level 2 universities, 25 to Level 3 universities, and 30 to Level 4 universities. We focus on career JF-equivalent pages and the career weighted number of articles in top-3 finance journals. Career records are defined to include the relevant research productivity measures up through one year after the individual moves to the new affiliation. This is to account for the forthcoming articles when he/she interviews for the new position. These career records are compiled from the Heck's Finance Literature Index (before 1990) and the data set that generates Table II. Panel A of Table V shows that, on average, the up-movers authored an average of 62.45 career JF-equivalent pages in the 16 finance journals, with a maximum of 504.11 pages and a minimum of 3.91 pages. The group, however, authored only a weighte d 1.44 career top-3 finance journal articles. The average teaching period before the change of affiliation is 8.53 years with a standard deviation of 4.98 years.

Panel B gives a more detailed breakdown of the statistics. For example, the average JF-equivalent pages authored by the individuals who moved from a Level 3 institution to a Level 2 institution is 96.37 and the average number of top-3 finance journal articles is 2.68. On the other hand, the same statistics are 33.86 JF-equivalent pages and 0.34 top-3 finance journal articles for those who moved from Level 5 to Level 4 institutions. Both statistics show that the research output is more demanding as one moves from a lower ranked institution to a higher ranked institution. In other words, it generally takes more human capital to move to a higher ranked institution. The research requirement, however, is the highest for those who moved from Level 3 to Level 1 institutions. In most cases, it takes more absolute effort to move two ranking classes upward than to move one ranking class upward even if the destination rank is the same. For example, it takes more effort to move from a Level 3 institution to a Level 1 in stitution than from a Level 2 institution to a Level 1 institution. One of the exceptions is the case from Level 4 to Level 1 institution. This case only contains two observations although it represents a giant leap forward. Examining the data we find these two observations are each associated with a short employment history before the relocation. Therefore, other factors of human capital endowment such as the Ph.D. granting institution and the average publication per year probably play important roles in these instances.

In further comparing statistics reported in Panel B of Table V with Panel B of Table IV, we observe a very interesting phenomenon. It takes, on average, higher research productivity for a faculty member to move to a certain level institution than another faculty that stays at the same institution of the same rank. For example, the average JF-equivalent pages for a Level 1 institution is 60.55 pages (Table IV, Panel B) and it takes much more effort in order to be accepted into such an institution from a lower level one. As a matter of fact, it takes an average of 104.22 (106.88) pages in order to move from a Level 3 (Level 2) institution to a Level 1 institution (Table V, Panel B). This represents a nearly doubling of the research productivity relative to existing faculty at the destination university. This finding is not limited to Level 1 institutions; the same conclusions apply to all ranks of institutions. The higher threshold obviously deters many who wish to move upward and that potentially explains why there are only a limited number of individuals who are able to move from a lower ranked institution to a higher level institution. An entrenchment effect induced by the tenure system offers a potential explanation for this finding.

To further examine the relationship between a career move-up and human capital, we estimate an ordered logit model. The results are presented in Table VI. The dependent variable in the proportional ordered logit model is 1, 2, 3 or 4 that correspond to the ranking of the destination university. Several explanatory variables serve as proxies for a researcher's endowment of human capital including the individual's total career JF-equivalent page counts, the career per year JF-equivalent page counts, the career total weighted number of articles published in the top-3 journals, the rank of the Ph.D. granting institution (according to our ranking), the number of years of teaching experience, and a year dummy (d) equal to 0 if an individual moved during 1990-1995 and equal to 1 otherwise. We use alternative research productivity variables in the four different empirical models presented in Table VI.

In Model 1, we use the total career JF-equivalent pages as a measurement of research productivity. As expected, the research productivity variable is positive and statistically significant at the 1% level. The Ph.D. granting institution variable is measured as 1 for the highest quality and 5 for the lowest quality. Consequently, the statistically significant and negative sign on the parameter suggests that a higher-ranked (lower ordered) Ph.D. granting institution does increase the odds of an upward move. Given the fact that the average teaching experience of the group is 8.5 years, the type of Ph.D. granting institution has a long-lasting effect on employment. In other words, the rank of the Ph.D. granting institution carries a value beyond the initial placement effect. The year dummy is not significant, meaning that labor market conditions do not significantly change the odds of moving up to a higher ranked institution given a level of human capital. Interestingly, the parameter estimate on the years of te aching experience is negative and statistically significant at the 10% level. The negative sign suggests that the longer a faculty has been away from the environment of the Ph.D. program, the more his or her human capital depreciates. Since the years of teaching experience is expected to be highly correlated with a researcher's age, the negative effect could potentially be an age effect. However, the lack of this personal information precludes us from separating these two effects. The model's predictive ability, measured by the rank correlation between observed responses and predicted probabilities, is 77.8%.

Model 2 uses JF-equivalent page counts per year as a measurement of research productivity. This variable also carries a positive sign and is significant at the 1% level. Conclusions on all other variables remain the same except that now the teaching experience variable is not significant. Models 3 and 4 are similar to Models 1 and 2 except that, in these models we add another variable to measure research productivity, namely, the total weighted number of articles in the top-3 finance journals. In Model 3, both measures of research productivity while carrying positive signs are statistically insignificant. We suspect this the result of multicollinearity in the regressors. The correlation analysis between the JF-equivalent pages and the weighed number of top-3 articles shows a simple correlation coefficient of 0.88. Moreover, although not individually significant, the two productivity measures are jointly significant at the 1% level. We further address the multicollinearity issue in Model 4.

Model 4 includes both per year JF-equivalent pages and top-3 finance journal articles as a measurement of research productivity. These two research productivity measures are less correlated (r = 0.46) than the two used in Model 3, and thus lessen the multicollinearity problem. The parameter on the variable for the top-3 finance journal articles is significant at the 5% level and the per year JF-equivalent page variable is significant at the 10% level. This result is more consistent with the results found in Models 1 and 2 in which a researcher's career-long record is as important as his/her average annual productivity.

IV. Conclusions

The purpose of this study is two-fold. In the first part of the paper, we provide an institutional ranking of research productivity in finance on a global scale. In the second part of the paper, we analyze several perspectives on an upward career move in the labor market in finance. We manually collect all publication records from a set of 16 core finance journals during the period from 1990 to 2001. Our major findings are: 1) Publications in the core finance journals are highly concentrated in a few institutions. For example, the first eleven institutions account for 20% of all publications while the first 37 institutions (out of 923 total institutions) account for 40% of all publications. 2) North American institutions (mainly US institutions) are the major contributors to the finance literature. The top twenty institutions are entirely US institutions. 3) Comparing the first and the second half of the sample period, we find many of the most-improved-institutions are outside of North America. This suggests that research in finance is gaining momentum in Europe and Asia, possibly responding to the growth in the financial markets in those regions. 4) Labor turnover is quite low for the upward movement market. Among a total of 4,990 academic authors, we are able to confirm only 88 individuals who successfully moved to a higher-level institution during the entire 12-year period. 5) Because of the higher threshold employed by many institutions for recruiting non-fresh Ph.D.s, an individual who is able to relocate to a higher-ranked institution must exhibit a research record that is approximately two times stronger than that of an average existing faculty member in the destination institution. We interpret this finding as a potential entrenchment effect induced by the tenure system. 6) The ability to move to a higher ranked institution is positively related to the total JF-equivalent page counts, the number of articles published in the top-3 finance journals, per year JF-equivalent page counts, and the Ph.D. granting institution rank. There is some evidence, however, that the ability to move to a higher ranked institution is negatively related to the years of teaching experience.

Potential caveats in our database and findings are: 1) it is possible that a small number of authors from other disciplines may have contributed to the finance literature. As a consequence, this may overstate the rankings of certain finance departments; 2) some individuals with strong background in economics also publish finance research in elite economic journals. Since these individuals are disproportionately affiliated with top-ranked universities, excluding these publications is likely to understate the total research productivity of these institutions; and 3) many in our profession would consider a distinguished professorship at an equal or emerging lower level school to also represent an upward move. Excluding these cases reduces our sample size used to assess labor mobility.

[FIGURE 1 OMITTED]
Table 1. Summary Statistics of the Research Productivity in a
Set of 16 Core Finance Journals from 1990 to 2001

This table contains some preliminary summary statistics of the
research productivity that bases on a set of 16 core journals.
These journals are Journal of Finance, Journal of Financial
Economics, Review of Financial Studies, Journal of Business,
Journal of Financial and Quantitative Analysis, Financial
Management, Journal of Financial Research, Journal of Banking
and Finance, Journal of Portfolio Management, Financial Analysts
Journal, Financial Review, Journal of Financial Intermediation,
Journal of Financial Services Research, Journal of Futures Markets,
Journal of International Money and Finance, and Journal of Business
Finance and Accounting. There are 4,990 academic authors in 923
academic institutions that published in the 16 core finance
journals. Panel C suggests that there are only about 14% of the
academic authors who have published at least five articles
(unweighted) or more. The "sum" statistics are not identical in
Panel and Panel B because: 1) some authors may have both academic
and non-academic affiliations, and 2) some authors may have more
than one academic affiliations.


Panel A. By Academic Institutions

 Weighted No. of Unweighted No. JF-equivalent
 Articles of Articles Page Counts

Mean 7.08 14.38 130.52
Median 1.50 3.00 25.11
Standard Deviation 14.38 29.08 301.81
Kurtosis 33.73 38.16 40.68
Skewness 4.74 4.92 5.43
Minimum 0.08 1.00 1.00
Maximum 172.06 370.00 3521.86
Sum 6,531.70 13,272 120,467.92
Count 923 923 923

Panel B. By Academic Authors

 Weighted No. of Unweighted No. JF-equivalent
 Articles of Articles Page Counts

Mean 1.31 2.53 24.26
Median 0.83 1 13.03
Standard Deviation 1.54 2.86 31.58
Kurtosis 18.61 18.15 22.08
Skewness 3.52 3.52 3.82
Minimum 0.20 1 0.65
Maximum 16.50 34 392.22
Sum 6,536.35 12,601 121,055.42
Number of authors 4,990 4,990 4,990

Panel C. Number of Unweighted Publications for All Authors in Academic
Institutions

No. of Unweighted Publications
(1990-2001) No. of Authors % of Total Cumulative %

One publication 2,719 54.49
Two publications 856 17.15 71.64
Three publications 437 8.76 80.40
Tour publications 260 5.21 85.61
Five publications 186 3.73 89.34
Six publications 128 2.57 91.90
Seven publications 101 2.02 93.93
Eight publications 87 1.74 95.67
Nine publications 57 1.14 96.81
Ten to twelve publications 87 1.74 98.56
Thirteen or more 72 1.44 100.00
Total 4,990 100.00

Table II

Top 100 World Research Productivity Ranking of Academic Institutions
(1990-2001) Based upon Journal of Finance-equivalent Pages Published in
a Set of 16 Finance Journals

This table presents the top 100 most productive institutions in the
world. The ranking in Column (1) is based on the weighted JF-equivalent
pages in 1990-2001 shown in Column (3). The JF-equivalent pages are
weighted by coauthorship and coaffiliations of the authors. While there
are universities from other countries (as specified in Column 2), the
majority of the schools are from the North America (US and Canada).
Columns (4) and (6) rank the universities based on the JF-equivalent
pages in 1990-95 and 1996-2001. The absolute changes in JF-equivalent
pages is in Column (8). Column (9) provides an improvement ranking based
on the magnitude of absolute changes in JF-pages.

 (1) (2) (3)


Overall Affiliation
Ranking (Countries if JF-pages
(1990-2001) other than US) (1990-2001)

1 NYU 3521.86
2 U of Penn 3179.40
3 U of Chicago 2343.71
4 Harvard U 2260.55
5 U of Michigan 2117.43
6 UCLA 1984.69
7 Columbia U 1801.16
8 Northwestern U 1561.61
9 Ohio State U 1557.12
10 Duke U 1543.13
11 Cornell U 1482.96
12 U of Illinois 1445.88
13 Stanford U 1347.52
14 U of Rochester 1222.36
15 MIT 1145.63
16 Indiana U 1128.72
17 U of Southern California 1126.57
18 UC-Berkeley 1097.39
19 Boston College 1029.80
20 U of Texas-Austin 1012.76
21 Virginia Tech 1004.73
22 London Business School (UK) 955.71
23 U of Florida 946.53
24 U of British Columbia (Canada) 940.18
25 Purdue U 922.60
26 Rutgers U 905.27
27 U of North Carolina 873.49
28 Vanderbilt U 871.79
29 U of Washington 843.26
30 Yale U 824.30
31 Arizona State U 823.10
32 U of Maryland 788.28
33 U of Georgia 753.65
34 Carnegie Mellon U 701.38
35 UW-Madison 699.37
36 SMU 676.68
37 U of Notre Dame 674.17
38 Baruch College 666.65
39 Georgetown U 665.73
40 Hong Kong University of Science
 and Technology (Hong Kong) 657.43
41 Penn State U 650.49
42 Emroy U 635.90
43 City U Business School (UK) 625.10
44 Georgia State U 611.19
45 U of Houston 610.86
46 Washington U 610.18
47 Michigan State U 601.68
48 U of Iowa 596.79
49 Florida Atlantic U 594.95
50 Texas A&M U 589.36
51 U of Missouri 576.08
52 Southern Illinois U 565.08
53 U of Virginia 558.71
54 LSU 553.94
55 U of Miami 549.67
56 U of Oklahoma 518.47
57 Iowa State U 515.33
58 Lancaster U (UK) 513.68
59 U of Strathclyde (UK) 513.48
60 Dartmouth College 508.57
61 U of Minnesota 494.92
62 Fordham U 491.61
63 Santa Clara U 486.78
64 Florida State U 486.64
65 Boston U 477.27
66 U of South Carolina 476.97
67 Chinese University of Hong Kong
 (Hong Kong) 472.13
68 U of Manchester (UK) 467.74
69 U of Alabama 463.97
70 U of New South Wales (Australia) 456.98
71 Wilfrid Laurier U (Canada) 445.72
72 Rice U 438.94
73 U of Utah 430.36
74 UC-Irvine 424.51
75 Tulane U 416.43
76 UC-Davis 414.41
77 London School of Economics (UK) 405.81
78 Loyola U-Chicago 403.18
79 Bentley College 401.15
80 Princeton U 396.57
81 U of Arizona 396.55
82 U of Toronto (Canada) 394.49
83 U of Alberta (Canada) 393.65
84 U of Kansas 388.36
85 SUNY-Binghamton 385.37
86 National U of Singapore (Singapore) 383.85
87 Clemson U 380.50
88 Temple U 377.88
89 Northern Illinois U 376.73
90 McGill U (Canada) 364.42
91 Tel-Aviv U (Israel) 364.24
92 HEC (France) 363.65
93 INSEAD (France) 356.41
94 North Carolina State U 348.24
95 Hebrew U (Israel) 347.31
96 Simon Fraser U (Canada) 337.25
97 U of Tennessee 332.68
98 Case Western Reserve U 332.61
99 U of colorado 331.25
100 Erasmus U (Netherlands) 327.20

 (1) (4) (5) (6) (7)


Overall
Ranking Ranking for JF-pages Ranking for JF-pages
(1990-2001) (1990-95) (1990-95) (1996-2001) (1996-2001)

1 1 1610.89 1 1910.97
2 2 1407.12 2 1772.29
3 3 1241.12 4 1102.59
4 5 1085.21 3 1175.34
5 4 1232.35 7 885.08
6 6 941.85 5 1042.84
7 10 855.86 6 945.30
8 9 859.48 10 702.13
9 7 866.40 11 690.72
10 8 864.11 12 679.02
11 12 720.26 9 762.71
12 11 834.03 16 611.85
13 20 521.97 8 825.55
14 13 688.35 25 534.01
15 21 513.90 15 631.73
16 19 527.31 17 601.41
17 17 562.68 20 563.90
18 18 555.77 22 541.62
19 16 606.87 34 422.93
20 32 367.83 13 644.93
21 15 625.12 41 379.61
22 35 311.35 14 644.36
23 26 412.32 24 534.21
24 14 668.46 66 271.72
25 31 380.02 21 542.58
26 23 439.14 28 466.14
27 33 363.95 26 509.54
28 25 420.08 31 451.71
29 29 390.10 30 453.16
30 57 234.33 18 589.97
31 22 460.32 45 362.79
32 48 251.70 23 536.58
33 28 404.53 48 349.12
34 37 302.39 37 399.00
35 24 437.99 69 261.38
36 34 314.96 46 361.72
37 53 241.57 32 432.60
38 56 235.32 33 431.32
39 87 170.82 27 494.91
40
 147 85.80 19 571.63
41 75 193.48 29 457.01
42 54 241.22 38 394.68
43 46 258.03 44 367.07
44 40 293.77 53 317.42
45 49 251.22 47 359.64
46 58 233.49 42 376.70
47 63 220.14 40 381.54
48 73 197.46 36 399.33
49 82 181.88 35 413.07
50 39 299.37 59 290.00
51 60 228.59 49 347.49
52 43 274.55 58 290.53
53 88 167.55 39 391.16
54 30 382.15 109 171.79
55 44 271.42 64 278.25
56 55 237.07 62 281.40
57 42 283.99 78 231.34
58 84 178.13 51 335.56
59 41 284.98 79 228.49
60 69 206.49 56 302.08
61 66 212.82 61 282.10
62 51 243.35 72 248.25
63 90 166.54 52 320.24
64 38 300.67 99 185.97
65 27 408.78 222 68.49
66 76 192.36 60 284.61
67
 131 100.05 43 372.08
68 85 172.81 57 294.94
69 36 308.03 120 155.94
70 65 213.65 75 243.33
71 50 247.71 91 198.01
72 109 136.73 55 302.22
73 92 164.19 68 266.17
74 59 229.39 95 195.12
75 125 112.27 54 304.16
76 172 69.04 50 345.37
77 99 153.05 71 252.76
78 52 241.89 117 161.29
79 97 154.47 73 246.68
80 91 164.29 77 232.28
81 77 191.60 89 204.94
82 71 204.54 97 189.95
83 121 115.29 63 278.36
84 74 196.56 96 191.81
85 83 178.93 87 206.44
86 108 137.49 74 246.36
87 45 259.51 149 120.99
88 67 211.52 113 166.37
89 102 150.20 81 226.53
90 78 190.69 106 173.74
91 79 185.70 104 178.54
92 62 221.63 131 142.01
93 80 185.14 110 171.27
94 61 228.11 152 120.13
95 89 166.76 103 180.55
96 123 115.24 82 222.01
97 47 256.78 210 75.89
98 141 91.77 76 240.83
99 119 119.86 86 211.38
100 132 99.81 80 227.39

 (1) (8) (9)
 Ranking for
 Improvement
Overall Absolute Based on
Ranking Changes in Absolute
(1990-2001) JF-pages Changes

1 300.08 7
2 365.17 2
3 -138.53 86
4 90.13 46
5 -347.27 99
6 100.98 39
7 89.44 47
8 -157.35 89
9 -175.68 90
10 -185.09 94
11 42.45 55
12 -222.18 96
13 303.58 6
14 -154.34 88
15 117.82 33
16 74.10 49
17 1.22 66
18 -14.16 71
19 -183.94 93
20 277.11 9
21 -245.51 97
22 333.01 4
23 121.89 31
24 -396.74 100
25 162.56 21
26 27.00 59
27 145.59 27
28 31.63 56
29 63.06 52
30 355.64 3
31 -97.53 82
32 284.87 8
33 -55.41 78
34 96.61 41
35 -176.62 91
36 46.76 53
37 191.03 18
38 196.00 16
39 324.09 5
40
 485.83 1
41 263.52 12
42 153.47 25
43 109.04 34
44 23.65 60
45 108.42 36
46 143.21 28
47 161.40 22
48 201.87 15
49 231.20 13
50 -9.37 69
51 118.89 32
52 15.97 61
53 223.61 14
54 -210.36 95
55 6.83 64
56 44.33 54
57 -52.66 77
58 157.43 23
59 -56.49 79
60 95.60 42
61 69.28 50
62 4.90 65
63 153.70 24
64 -114.71 84
65 -340.29 98
66 92.25 43
67
 272.03 11
68 122.13 30
69 -152.10 87
70 29.68 57
71 -49.70 76
72 165.49 19
73 101.98 38
74 -34.28 74
75 191.89 17
76 276.33 10
77 99.72 40
78 -80.59 81
79 92.21 44
80 67.99 51
81 13.34 63
82 -14.59 72
83 163.07 20
84 -4.75 67
85 27.51 58
86 108.87 35
87 -138.52 85
88 -45.15 75
89 76.33 48
90 -16.95 73
91 -7.16 68
92 -79.62 80
93 -13.87 70
94 -107.98 83
95 13.78 62
96 106.77 37
97 -180.89 92
98 149.06 26
99 91.52 45
100 127.58 29

Table III

World Research Productivity (1990-2001) by JF-equivalent
Pages in Top-3 Finance Journals (Journal of Finance, Journal
of Financial Economics, and Review of Financial Studies)

This table provides a ranking of world research productivity
by JF-equivalent pages in top-3 finance journals only. There are
319 academic institutions that publish at least one unweighted
articles in the top three finance journals. In other words, there
are 604 academic institutions that did not publish any articles in
the top three journals. Among the 319 academic institutions that
published at least one unweighted article in top-3 finance journals,
the rank correlation coefficient between JF-equivalent pages in the
top-3 and in the 16 finance journals is 81.26% and the rank correlation
coefficient between the number of weighted articles in the top-3 and the
16 finance journals is 80.54%. While there are universities from other
countries (as specified in Column 2), the majority of the schools are
from the North America (US and Canada). Columns (1), (4), and (6) rank
the universities based on the JF-equivalent pages in 1990-2001, 1990-95,
and 1996-2001, respectively. The absolute changes in JF-equivalent pages
are in Column 8. Column 9 provides an improvement ranking based on the
magnitude of absolute changes in JF-pages.



(1) (2) (3)


Overall Affiliation
Ranking (Countries if JF-pages
(1990-2001) other than US) (1990-2001)

1 U of Penn 2059.95
2 NYU 2013.63
3 U of Chicago 1899.91
4 Harvard U 1601.52
5 U of Michigan 1426.49
6 UCLA 1285.68
7 Columbia U 1076.87
8 Northwestern U 1061.97
9 Duke U 1056.12
10 Ohio State U 1006.44
11 U of Rochester 1002.84
12 Stanford U 973.45
13 U of Southern California 837.19
14 MIT 812.03
15 Cornell U 769.63
16 U of Illinois 619.45
17 Carnegie Mellon U 618.42
18 U of British Columbia (Canada) 615.43
19 UC-Berkeley 593.01
20 U of North Carolina 568.59
21 Arizona State U 566.56
22 Boston College 560.50
23 U of Florida 499.69
24 London Business School (UK) 489.97
25 U of Maryland 472.21
26 Purdue U 462.93
27 U of Texas-Austin 457.16
28 Yale U 455.34
29 Vanderbilt U 434.12
30 U of Washington 423.63
31 Washington U 383.94
32 Penn State U 358.78
33 Indiana U 358.07
34 U of Iowa 347.82
35 Virginia Tech 347.79
36 SMU 341.13
37 UW-Madison 332.57
38 U of Georgia 320.28
39 UC-Davis 308.62
40 U of Notre Dame 297.66
41 Dartmouth College 269.62
42 Princeton U 253.96
43 UC-Irvine 252.75
44 Tulane U 251.25
45 Georgetown U 242.53
46 U of Virginia 241.49
47 Hong Kong University of Science
 and Technology (Hong Kong) 234.14
48 U of Minnesota 231.84
49 U of Utah 221.40
50 Emory U 207.96
51 U of Arizona 195.01
52 U of Oregon 194.72
53 Rice U 187.12
54 Michigan State U 186.25
55 U of Colorado 173.78
56 Tel-Aviv U (Israel) 169.54
57 LSU 163.37
58 Georgia State U 156.53
59 INSEAD (France) 142.69
60 U of Toronto (Canada) 140.87
61 McGill U (Canada) 139.85
62 U of Missouri 139.44
63 Baruch College 138.18
64 U of Alberta (Canada) 134.03
65 HEC (France) 133.04
66 UC-Riverside 131.44
67 Iowa State U 125.96
68 U of Oklahoma 122.48
69 Rutgers U 122.39
70 Hebrew U (Israel) 113.51
71 U of Auckland (New Zealand) 111.63
72 UC-San Diego 111.52
73 U of Houston 109.66
74 U of Miami 109.52
75 Case Western Reserve U 103.86
76 U of Texas-Dallas 102.41
77 U of Pittsburgh 101.71
78 London School of Economics (UK) 101.69
79 U of Western Ontario (Canada) 101.06
80 Clemson U 96.35
81 Norwegian School of Management
 (Norway) 92.60
82 Simon Fraser U (Canada) 86.81
83 Santa Clara U 80.81
84 Queen's U (Canada) 80.77
85 SUNY-Buffalo 80.57
86 Texas A&M U 79.15
87 U of Rhode Island 76.80
88 U de Toulouse (France) 75.94
89 George Mason U 69.14
90 ESSEC (France) 66.24
91 Wayne State U 64.94
92 Bentley College 63.07
93 U of Bern (Germany) 62.76
94 U of Amsterdam (Netherlands) 62.10
95 Washington State U 61.19
96 Northeastern U 60.94
97 Loyola U-Chicago 60.39
98 California Institute of Technology 59.77
99 Chinese University of Hong Kong
 (Hong Kong) 58.72
100 College of William and Mary 58.14

(1) (4) (5) (6) (7)


Overall
Ranking Ranking for JF-pages Ranking for JF-pages
(1990-2001) (1990-95) (1990-95) (1996-2001) (1996-2001)

1 3 824.81 1 1235.14
2 4 791.64 2 1221.99
3 1 1044.23 3 855.68
4 5 773.46 4 828.06
5 2 836.56 8 589.92
6 6 596.86 5 688.82
7 11 450.37 7 626.50
8 7 558.77 10 503.20
9 10 518.35 9 537.77
10 9 546.59 12 459.85
11 8 551.28 13 451.56
12 18 291.88 6 681.57
13 12 432.59 16 404.60
14 15 376.08 14 435.95
15 17 308.03 11 461.60
16 16 332.97 26 286.48
17 21 254.95 18 363.47
18 13 421.25 37 194.18
19 19 287.89 23 305.12
20 32 154.33 15 414.26
21 20 286.00 28 280.57
22 14 386.32 42 174.18
23 27 183.96 22 315.72
24 31 158.29 21 331.68
25 40 97.78 17 374.43
26 24 197.55 29 265.39
27 38 107.65 19 349.51
28 35 120.39 20 334.95
29 34 133.95 25 300.17
30 26 184.77 30 238.86
31 25 191.82 38 192.12
32 47 76.13 27 282.65
33 33 138.56 33 219.51
34 36 118.82 31 229.00
35 22 207.00 48 140.80
36 30 160.68 41 180.45
37 23 200.93 52 131.64
38 28 180.96 50 139.32
39 208 4.00 24 304.62
40 46 76.48 32 221.17
41 44 80.63 40 188.99
42 41 96.38 44 168.32
43 29 171.59 64 81.15
44 54 60.45 39 190.80
45 67 46.67 35 195.87
46 66 46.98 36 194.51
47
 107 22.50 34 211.64
48 39 102.38 53 129.46
49 51 68.79 47 152.61
50 70 41.00 45 166.96
51 79 35.13 46 159.88
52 112 21.12 43 173.60
53 64 47.13 49 139.99
54 62 48.35 51 137.89
55 59 53.67 55 120.11
56 45 80.28 63 89.26
57 37 115.90 93 47.48
58 42 94.37 77 62.16
59 49 72.08 67 70.60
60 94 27.63 57 113.25
61 96 26.02 56 113.84
62 73 38.91 60 100.53
63 50 71.91 71 66.27
64 191 6.00 54 128.03
65 43 93.33 102 39.71
66 74 37.00 61 94.44
67 53 61.75 73 64.21
68 48 73.21 86 49.28
69 58 54.99 70 67.41
70 68 44.87 69 68.64
71 82 33.10 65 78.53
72 52 63.52 91 48.00
73 72 39.00 66 70.66
74 199 4.67 58 104.85
75 71 40.67 76 63.19
76 n/a n/a 59 102.41
77 69 44.11 81 57.60
78 61 50.52 84 51.17
79 75 36.86 74 64.20
80 60 50.97 95 45.38
81
 n/a n/a 62 92.60
82 105 22.87 75 63.94
83 152 11.67 68 69.15
84 56 56.82 131 23.95
85 77 36.00 97 44.57
86 55 59.98 150 19.16
87 57 55.46 135 21.34
88 125 14.76 78 61.18
89 88 30.00 103 39.14
90 n/a n/a 72 66.24
91 210 4.00 79 60.94
92 119 17.42 94 45.65
93 63 47.95 166 14.81
94 133 13.16 87 48.94
95 113 21.12 101 40.07
96 65 47.00 169 13.94
97 182 6.67 83 53.72
98 101 24.20 109 35.57
99
 159 10.45 89 48.27
100 n/a n/a 80 58.14

(1) (8) (9)
 Ranking for
 Improvement
Overall Absolute Based on
Ranking Changes in Absolute
(1990-2001) JF-pages Changes

1 410.32 2
2 430.34 1
3 -188.55 97
4 54.60 47
5 -246.64 100
6 91.96 31
7 176.13 11
8 -55.58 90
9 19.42 62
10 -86.75 94
11 -99.72 96
12 389.69 3
13 -27.99 80
14 59.88 42
15 153.57 14
16 -46.49 88
17 108.52 25
18 -227.07 99
19 17.22 64
20 259.93 6
21 -5.43 75
22 -212.14 98
23 131.76 19
24 173.38 12
25 276.65 5
26 67.84 38
27 241.86 7
28 214.56 8
29 166.22 13
30 54.10 48
31 0.30 73
32 206.52 9
33 80.96 36
34 110.18 24
35 -66.20 91
36 19.77 61
37 -69.29 93
38 -41.64 87
39 300.62 4
40 144.69 18
41 108.37 26
42 71.94 37
43 -90.44 95
44 130.35 20
45 149.20 16
46 147.54 17
47
 189.14 10
48 27.08 58
49 83.82 35
50 125.96 21
51 124.75 22
52 152.48 15
53 92.86 29
54 89.54 32
55 66.44 39
56 8.97 69
57 -68.42 92
58 -32.21 81
59 -1.48 74
60 85.62 34
61 87.82 33
62 61.62 41
63 -5.64 77
64 122.03 23
65 -53.63 89
66 57.44 45
67 2.46 71
68 -23.93 79
69 12.42 66
70 23.77 59
71 45.43 51
72 -15.52 78
73 31.66 55
74 100.18 28
75 22.52 60
76 102.41 27
77 13.49 65
78 0.65 72
79 27.34 57
80 -5.59 76
81
 92.60 30
82 41.07 52
83 57.48 44
84 -32.88 82
85 8.57 70
86 -40.82 86
87 -34.12 85
88 46.43 50
89 9.14 68
90 66.24 40
91 56.94 46
92 28.22 56
93 -33.14 84
94 35.77 54
95 18.95 63
96 -33.06 83
97 47.05 49
98 11.37 67
99
 37.81 53
100 58.14 43

Table IV. Summary Statistics of Changing Academic Affiliations

This table presents some summary statistics of changing academic
affiliations. Academic institutions are classified into five
levels. These five classes are based on the cumulative percentile
of the JF-equivalent pages in Table II. Essentially, the top 20%
of the institutions in Table II are considered to be Level 1.
The next 20% of the institutions are considered to be in Level 2.
Levels 3, 4, and 5 are similarly defined. If an author made multiple
affiliation changes over the period of 1990-2001, we counted him or
her multiple times. Hence, the total (5,268) is larger than the 4,990
reported in Table I. In Panel A, "equal level" means the author move
to a school that has the same level as his or her original school.
Similarly, a "higher" or "lower level" school has the same
interpretation.

Panel A. Descriptive Staatistics


Type of Affiliation Wt. Articles
Changes (1999-2001) Freq. Avg. Std Dev.

Stay at the same institutions 1,549 1.95 1.55
Change to the institution of
equal level 254 2.82 2.41
Change to a lower level
institution 384 3.01 2.26
Change to a higher level
institution 362 2.94 2.16
Only one publication (only
published one articles and
cannot determine the
affiliation change) 2,719 0.52 0.23
Total 5,268

 JF-equivalent
Type of Affiliation Unwt. Articles Pages
Changes (1999-2001) Avg. Std Dev. Avg.

Stay at the same institutions 3.76 2.81 35.71
Change to the institution of
equal level 5.67 4.70 52.66
Change to a lower level
institution 5.91 4.18 57.74
Change to a higher level
institution 5.67 4.13 54.26
Only one publication (only
published one articles and
cannot determine the
affiliation change) 1 N/A 9.44
Total

 JF-equival
 ent
Type of Affiliation Pages
Changes (1999-2001) Std Dev.

Stay at the same institutions 32.08
Change to the institution of
equal level 50.28
Change to a lower level
institution 48.42
Change to a higher level
institution 47.05
Only one publication (only
published one articles and
cannot determine the
affiliation change) 6.16
Total

Panel B. Productivity for those Who Stay at the Same Institution

Stay at the
Same Institutions Wt. Articles Unwt. Articles
(1999-01) Freq. Avg. Std Dev. Avg. Std Dev.

Level 1 204 2.65 2.31 4.76 4.15
Level 2 281 2.10 1.52 3.98 2.72
Level 3 314 2.16 1.74 4.15 3.30
Level 4 406 1.83 1.15 3.70 2.18
Level 5 345 1.38 0.87 2.69 1.34
Total 1,549

Stay at the JF-equivalent
Same Institutions Pages
(1999-01) Avg. Std Dev.

Level 1 60.55 52.07
Level 2 43.62 34.78
Level 3 37.30 28.45
Level 4 29.43 19.13
Level 5 20.51 13.48
Total

Table V

Statistics for Individuals Who Move to a Higher Ranked Institution

This table reports descriptive statistics on research productivity for a
sample of 88 career move-up faculty. Five institutional rankings are
defined based upon the research productivity measures reported in Table
I. Essentially, the top 20% of the institutions in Table II are
considered to be Level 1. The next 20% of the institutions are
considered to be in Level 2. Levels 3, 4, and 5 are similarly defined.
Years of teaching experience is measured by the time span between
receiving a Ph.D. degree and moving to a different institution. Some
individuals may begin their teaching before receiving a Ph.D., but still
counted as one- year teaching experience. The career page count includes
the lifetime productivity of the faculty member up through one year
after the move.

Panel A. Summary Statistics of Career Move-Up Faculty

 Career Total
 No. of JF-
 No. of Career equivalent Page
 Weighted JF, JFE, Counts from 16
 and RFS Articles Journals (up to
 (up to 1 Yr after 1 Yr after the
 the Move) Move)

Mean 1.44 62.45
Median 0.59 42.61
Standard Deviation 2.04 66.96
Kurtosis 6.16 21.40
Skewness 2.35 3.78
Minimum 0.00 3.91
Maximum 10.16 504.11
No. of Faculty 88 88



 Per Yr Career Avg.
 of JF-equivalent
 Page Counts (up to Yrs of Teaching
 1 Yr after the Experience when
 Move) Moved

Mean 8.16 8.53
Median 6.32 8.00
Standard Deviation 6.80 4.98
Kurtosis 3.25 1.39
Skewness 1.71 1.11
Minimum 0.44 1
Maximum 31.64 26
No. of Faculty 88 88

Panel B. A Detail Breakdown of Career Move-Up Faculty

Type of Move Career JF-equivalent Pages in
(from a lower 16 Finance Journals up to 1 Yr
level to a higher No. of after the Move
level university) Faculty Mean Std Dev.

From 2 to 1 11 104.22 140.75
From 3 to 1 3 106.88 64.02
From 4 to 1 2 44.51 8.84
From 3 to 2 13 96.37 64.72
From 4 to 2 4 70.96 33.26
From 4 to 3 17 55.19 44.71
From 5 to 3 8 56.13 37.92
From 5 to 4 30 33.86 27.12

Type of Move Career Total No. of
(from a lower Weighted JF, JEF, and
level to a higher RFS Articles
level university) Mean Std Dev.

From 2 to 1 2.79 3.10
From 3 to 1 3.77 2.86
From 4 to 1 1.17 0.70
From 3 to 2 2.68 2.80
From 4 to 2 1.40 0.79
From 4 to 3 1.31 1.31
From 5 to 3 1.22 1.87
From 5 to 4 0.34 0.51

Table VI. An Ordered Logistic Model of Finance
Faculty Upward Mobility

This table presents an ordered logistic model of
finance faculty upward mobility. Generally, a positive
coefficient means that a high value of the explanatory
variable will improve the odds of moving from a lower
ranked university to a higher ranked university. Because
higher ranked Ph.D. granting institution is ordered lower,
a negative coefficient actually improves the odds. The results
suggest that a higher research productivity and a better Ph.D.
granting institution contribute to higher odds in moving to a
higher level institutions and the larger number of years of
post-Ph.D. teaching experience contributes less odds in moving
to a better institution. Wald [chi square] statistics are in
the parentheses.



 Dependent Variable: Move Up
 Destinations
 (1, 2, 3, or 4); with 1 as the
 Highest Level
 Model 1 Model 2
Explanatory Estimated Estimated
Variables Coefficient Coefficient

Intercept 1 -0.3118 -1.6056
 (0.1976) (3.5436) *
Intercept 2 0.9932 -0.3256
 (2.0777) (0.1576)
Intercept 3 2.6272 1.2646
 (12.5405) *** (2.2875)
Total Career JF-equivalent Pages 0.0186
 (up to 1 Yr. after the Move) (11.3882) ***
Per Yr. Career JF-equivalent Pages 0.1240
 since receiving Ph.D. (up to 1 Yr. (11.4079) ***
 after the Move)
Total Career Weighted No. of Articles
 in JF, JFE or RFS (up to 1 Yr. after
 the Move)
Ph.D. Granting Institution Ranking -0.7813 -0.7584
 (from ito 5; with 1 as the Highest (14.4117) *** (13.8457) ***
 Level)
Year dummy (for 1990-95, d=0; for -0.3549 -0.1837
 1996-01, d=1) (2.2375) (0.1854)
 Year of Teaching Experience when -0.0945 0.0599
 the Faculty Moves (3.5995) * (1.8539)
-2 log Likelihood Function 200.97 202.06
Score Test for the Proportional Odds 9.73 3.66
 Assumption ([chi square])
Model Predictive Ability 77.8% 75.5%
No. of Observations 88 88

 Dependent Variable: Move Up
 Destinations
 (1, 2, 3, or 4); with 1 as the
 Highest Level
 Model 3 Model 4
Explanatory Estimated Estimated
Variables Coefficient Coefficient

Intercept 1 -0.3951 -1.2672
 (0.3128) (2.1086)
Intercept 2 0.9368 0.0944
 (1.8432) (0.0125)
Intercept 3 2.5800 1.7510
 (12.0329) *** (4.0598) **
Total Career JF-equivalent Pages 0.0100
 (up to 1 Yr. after the Move) (1.1800)
Per Yr. Career JF-equivalent Pages 0.0728
 since receiving Ph.D. (up to 1 Yr. (2.9907) *
 after the Move)
Total Career Weighted No. of Articles 0.2732 0.3360
 in JF, JFE or RFS (up to 1 Yr. after (1.3301) (4.3442) **
 the Move)
Ph.D. Granting Institution Ranking -0.7352 -0.7043
 (from ito 5; with 1 as the Highest (12.4064) *** (11.6381) ***
 Level)
Year dummy (for 1990-95, d=0; for -0.2235 -0.0854
 1996-01, d=1) (0.2569) (0.0394)
 Year of Teaching Experience when -0.0935 -0.0223
 the Faculty Moves (3.5358) * (0.1687)
-2 log Likelihood Function 199.65 197.61
Score Test for the Proportional Odds 13.60 9.28
 Assumption ([chi square])
Model Predictive Ability 78.5% 78.0%
No. of Observations 88 88

*** Significant at the 0.01 level.

** Significant at the 0.05 level.

* Significant at the 0.10 level.


(1.) The set of 16 finance journals are: Journal of Finance (proceeding issues included). Journal of Financial Economics, Review of Financial studies, Journal of Business, Journal of Financial and Quantitative Analysis, Financial Management, Journal of Financial Research, Journal of Banking and Finance, Journal of Portfolio Management, Financial Analysts Journal, Financial Review, Journal of Financial Intermediation, Journal of Financial Services Research, Journal of Futures Markets. Journal of International Money and Finance, and Journal of Business Finance and Accounting.

(2.) While Borokhovich et al. (1995) include Journal of Money Credit, and Banking, we substitute it by the Journal of Financial Intermediation for the reason that the former publishes mostly monetary economics papers.

(3.) Financial Review changed to a different printing style and font size in 1998. We adjusted its total pages separately with respect to articles published before and after 1998.

(4.) An interesting question arises: What is the proportion of all finance faculty who have published at least one article during the twelve-year sample period? This statistic is difficult to estimate, because one has to know the number of finance academics in the whole world. Hasselback (1999) provides some guidance; however, this source focuses on the North American universities and, therefore, will severely underestimate the finance community in the other parts of the world. Using this source, we estimate a number of statistics based only upon North American institutions. Excluding the non-academic and non-North American authors, approximately 56% of the North American faculty has published at least one article during the 12 years period. If the same approximation rules apply, then one can infer that roughly 25% of the North American finance faculty published two or more articles during the 12-year period. These statistics could not be generalized for the global sample for two reasons: 1) we do not know th e true size of non-North American academics and 2) North American academics are more heavily published than non-North American academics.

(5.) Because the institutional affiliation is based upon the time when the research was published, it is possible that the research is credited to the author's old affiliation if the author relocates. Presumably, the author has incentive to notify the publisher for affiliation changes, but we could not rule out the possibility that this did not happen and the old affiliation still gets the credit, Of course, who should receive the credit is a philosophical issue. One could argue that old affiliation should receive the credit because the bulk of the work can be expected to have been conducted before the affiliation was changed.

(6.) As discussed in Section 1, the JF-equivalent page count is our preferred ranking method. However, the correlation coefficient between JF-equivalent page count ranking and weighted page count ranking is 96.7%.

(7.) We also examine research improvement based upon percentage changes in JF-equivalent pages. Though not reported, results are available upon request.

(8.) There are a few schools that make significant improvements over the sub-periods, but not included in Table II due to a relative low overall productivity. These schools include Hong Kong Polytechnic University (Hong Kong) that ranked 12th in the improvement ranking (from 0.00 to 269.74 JF-pages), Tilburg University (Netherlands) ranked 14th (from 28.96 to 275.26 JF-pages), and University of Exeter (UK) ranked 17th (from 38.87 to 258.39 JF-pages). Detailed disclosures can be found at the authors' website (http://faculty.sba.udayton.edu/chan).

(9.) We also plot the distribution of the top-3 journal research productivity. Similar to Figure 1, the distribution is equally skewed. That is, we find 1.55%, 4%, 8.4%, and 17.76% of the 319 institutions published 20%, 40%, 60%, and 80% of the total JF-equivalent pages in the top-3 journals. The same statistics for the 16 core journals are 1.2%, 4%, 9.08%, and 19.35%, respectively. The graph, therefore, is nearly identical to Figure 1.

(10.) This number is more than the 4,990 reported in Table I, because some individuals have mode more than one career move. If someone made career moves N times during 1990-2001, we count them N times in Table IV.

(11.) One can easily see that this to be true by glancing over the job-listing directory organized by the Financial Management Association International.

(12.) Many in the finance profession may consider a distinguished professorship at an equal or emerging lower level school to also represent an upward move. Excluding these cases reduce our sample size. However, it would he difficult to determine the nature of this type of job move without more detailed information and subjective judgment.

(13.) Since more than 60% of the movers are either lateral or downward movers, and the research productivity of these movers are no less impressive than the upward movers (Table IV), future studies that examine the human capital requirement for these two categories of movers are interesting in its own right. Because of the confounding motivations for relocation decision, however, it would be necessary to separate voluntary from involuntary movers. Presumably, involuntary downward movers relocate because of the lack of human capital given the rank of their employers. The question then is how to identify who is an involuntary mover. Although interesting, we leave this topic to future studies.

References

Alexander, J.C. and R.H. Mabry, 1994, "Relative Significance of Journals, Authors, and Articles Cited in Financial Research," Journal of Finance 49, 697-712.

Borokhovich, K.A., R.J. Bricker, K.R. Brunarski, and B.S. Simkins, 1995, "Finance Research Productivity and Influence," Journal of Finance 50, 1691-1717.

Borokhovich, K.A., R.J. Bricker, and B.J. Simkins, 2000, "An Analysis of Finance Journal Impact Factors," Journal of Finance 55, 1457-1469.

Brown, L.D., 1996, "Influential Accounting Articles, Individuals, Ph.D. Granting Institutions and Faculties: a Citational Analysis," Accounting Organizations and Society 21, 723-754.

Chung, K.H., R.A.K. Cox, and J.B. Mitchell, 2001, "Citation Pattern in the Finance Literature," Financial Management 30, 99-118.

Collins, J.T., R.G Cox, and V. Stango, 2000 "The Publishing Patterns of Recent Economics Ph.D. Recipients," Economic Inquiry 38, 3 58-367.

Kam C. Chan, Carl R. Chen, Thomas L. Steiner *

Conboy, M.E., R. Dusansky, D. Drukker, and A. Kildegaard, 1995, "The Productivity of Economics Departments in the US: Publications in the Core Journals," Journal of Economic Literature 33, 1966-1971.

Gomez-Mejia, L.R. and D.B. Balkin, 1992, "Determinants of Faculty Pay: An Agency Theory Perspective," Academy of Management Journal 35, 921-955.

Hasselback, J.R., 1999, The 1999/2000 Prentice Hall Guide to Finance Faculty, Prentice Hall. Heck, J.L., 1998, Finance Literature Index, McGraw-Hill.

Niemi, A.W.Jr., 1987, "Institutional Contributions to the Leading Finance Journals, 1975-1986: A Note," Journal of Finance 42, 1389-1397.

Scott, L.C. and P.M. Mitias, 1996, "Trends in Rankings of Economics Departments in the US: an Update," Economic Inquiry, 378-400.

Zivney, T.L. and W.J. Bertin, 1992, "Publish or Perish: What the Competition is Really Doing," Journal of Finance 47, 295-329.

Comments from Kee Chung. Campbell Harvey, Lemma Senbet and Alex Triantis (the Editors), an anonymous referee, and programming assistance from Christine Lai are gratefully acknowledged We are responsible for the remaining errors.

* Kam C. Chan and Thomas L. Steiner are Associate Professors of Finance at the University of Dayton. Carl R. Chen is the William J. Hoben Professor of Finance at the University of Dayton.
COPYRIGHT 2002 Financial Management Association
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2002 Gale, Cengage Learning. All rights reserved.

 Reader Opinion

Title:

Comment:



 

Article Details
Printer friendly Cite/link Email Feedback
Author:Chan, Kam C.; Chen, Carl R.; Steiner, Thomas L.
Publication:Financial Management
Date:Dec 22, 2002
Words:13223
Previous Article:Asset liquidity and the determinants of asset sales by poorly performing firms.
Next Article:Does financial liberalization reduce financing constraints?


Related Articles
Imperfect labor mobility and unemployment in LDCs: reply.
Business Accreditation: To Enhance Global Labor Standards.
An examination of occupational mobility among full-time workers: as workers approached mid-career in the late 1990s, they saw an increase in their...
HSBC launches pension funds aimed at Muslims.
The production of reality; essays and readings on social interaction, 4th ed.
The employability imperative; schooling for work as a national project.
World migration 2008; managing labour mobility in the evolving global economy.
Ethnic entrepreneurship: the distinct role of ties.

Terms of use | Copyright © 2014 Farlex, Inc. | Feedback | For webmasters