Relative significance of insurance and actuarial journals and articles: a citation analysis.
The importance of evaluating journal quality has been documented in the finance literature by Alexander and Mabry (1994) and Zivney and Reichenstein (1994). Assessing journal quality is important because (1) hiring, promotion, and tenure decisions often are based in large part on the quality of a faculty member's research; (2) faculty are interested in directing research to outlets where publication is most appropriate and beneficial; (3) an editor's knowledge of a journal's quality is important in evaluating future editing decisions; and (4) an institution's knowledge of journal quality is valuable when few resources are available for the purchasing of research journals.
The issue of limited library resources is particularly relevant to many academics who research and teach in the areas of insurance and actuarial science. Gardner and Schmit (1995), in a survey of collegiate risk management and insurance (RMI) education programs in the United States and Canada, find that 20 percent of the 60 schools that offer an undergraduate RMI program have 50 or fewer students enrolled in the program every year. In addition, the survey also states that 56 percent of the schools that report offering undergraduate RMI studies offer only one or two classes. It is unlikely that these universities have significant funds available to provide resources that generally serve only a small percentage of all students and faculty.
The purpose of this article is to provide an objective measure of the quality of a sample of insurance and actuarial journals based upon the number of journal citations appearing in thirteen insurance and actuarial journals and sixteen of the leading finance journals. Rankings also are provided based on the average relative "impact" that the insurance and actuarial journals' articles have had on other research. The study reveals the degree to which knowledge is communicated between insurance and actuarial journals by observing the frequency with which these journals are cited by one another. Finally, the most frequently cited articles from the sample insurance and actuarial journals are observed and reported.
In recent years, citation data have been used extensively in finance research to rank research journals, rank the research productivity of individual authors and institutions, and to evaluate interjournal communication. Alexander and Mabry (1994) and Zivney and Reichenstein (1994) rank finance journals according to the number of citations found in finance journals during a five-year and one-year period, respectively. Borokhovich et al. (1995) use citation data to evaluate the finance research productivity of 661 academic institutions during the years 1989 through 1993, adjusting for the number of faculty in an institution and the quality of the journals in which the faculty published research.
Borokhovich, Bricker, and Simkins (1994) evaluate the communication of knowledge between eight finance journals during the years 1990 and 1991 by observing the frequency with which each journal is cited by the other journals in the study. Their results reveal the degree of influence that each journal has had on research published in other journals. Finally, Schwert (1993) uses citation data to examine the influence that the Journal of Financial Economics had on other finance research during the years 1974 through 1991.
In an attempt to provide a ranking of journals (primarily insurance and actuarial journals) based upon the opinions of insurance academics, Outreville and Malouin (1985) survey members of the American Risk and Insurance Association (ARIA) regarding their familiarity with a number of journals and the members' perception of the value of having a paper published in each journal. The survey was comprehensive in that it included 255 journals from most mainstream business disciplines. Their results indicate that, of the top 30 journals, nine are insurance and actuarial journals. Although Outreville and Malouin report the results for the total number of survey participants, they also divide the survey participants into four subgroups and report the results for each subgroup.(1) The results reported for each subgroup vary considerably, signifying the subjective nature of the study.
In their study ranking the research productivity of authors, employing institutions and degree-granting institutions, Cox and Gustavson (1990) use the subjective quality measure derived by Outreville and Malouin to assess the research contributions made by the top insurance researchers. Chung and Puelz (1992) provide an alternative method of evaluating the relative performance of insurance researchers by "identifying an empirical regularity in the frequency distribution of risk and insurance researchers' article publications."(2)
Using survey data, McNamara and Kolbe (1996) rank sixteen journals, primarily insurance journals, based upon the opinions of the deans (or his or her representative) of 126 business schools. Finally, Baur, Zivney, and Wells (1996) analyze the subscription practices of universities with regard to insurance and actuarial periodicals. They provide information regarding the different subscription practices based upon a school's accreditation, insurance and actuarial course offerings, and doctoral degree-granting status. They also rank the top insurance and actuarial journals based upon the sample journals' subscription rates.
While Outreville and Malouin, McNamara and Kolbe, and Baur, Zivney, and Wells provide various measures of insurance and actuarial journal quality, citation analysis is arguably a better measure of an insurance or actuarial journal's contribution to insurance and actuarial research. With respect to insurance and actuarial journal quality, it is quite valuable to know the opinions of the members of ARIA (as in Outreville and Malouin), business school deans (as in McNamara and Kolbe) and those who make journal subscription decisions (as in Baur, Zivney, and Wells). However, only with citation analysis does the determination of an insurance or actuarial journal's quality lie squarely in the hands of those who would be most capable of assessing the journal's quality with regard to insurance and actuarial research - those who have conducted insurance and actuarial research.(3) With citation analysis, citations to articles and journals essentially serve as "votes" cast by the authors citing the research. The more research an individual has published, the more "votes" he or she casts in determining the relative contribution that previously published articles have made to the citing author's research.
Conspicuously absent from the insurance literature is the use of citation data to evaluate research or journal quality. Most finance studies that use citation data employ the Social Sciences Citation Index (SSCI) and the impact factor derived by the SSCI to measure journal quality or the degree of communication between journals (see Borokhovich et al., 1995; Borokhovich, Bricker, and Simkins, 1994; and Schwert, 1993). The SSCI impact factor is derived each year for most finance journals by totaling the number of citations to articles published in a journal for the previous two years and dividing that number by the number of articles published in the journal during the same two years. According to this methodology, the higher the impact factor, the greater the contribution a journal's articles have made to further research and, consequently, the higher the "quality" of the journal.
The primary reason that citation data have not been used to evaluate insurance literature is documented by both Cox and Gustavson (1990, p. 265) and Chung and Puelz (1992, p. 489). Cox and Gustavson observe that, of the top nine insurance and actuarial journals listed by Outreville and Malouin, only the Journal of Risk and Insurance (JRI) and Insurance: Mathematics and Economics (IME) are indexed by the SSCI. More recently, the SSCI has included the Geneva Papers on Risk and Insurance Theory (GPT) and the Journal of Risk and Uncertainty (JRU) among its list of indexed journals.(4) In addition to the lack of insurance and actuarial journals included in the SSCI, other problems associated with using the SSCI are documented by Alexander and Mabry (1994, p. 700).(5)
In a study of this nature, it is critical to convey exactly what the citation analysis measures. This study is based on citations found in thirteen insurance and actuarial journals and sixteen finance journals only to articles published in the thirteen insurance and actuarial journals. The sample journals are listed in the Appendix. It is important to point out that the study is limited to analyzing the relative significance of the thirteen sample insurance and actuarial journals and their articles. As a result, none of the sample finance journals will be found in the list of the most frequently cited journals. Also, frequently cited risk and insurance-related articles that have been published in many of the finance journals will not appear in the results as being frequently cited. Although these finance articles and their journals may be frequently cited by the sample insurance, actuarial, and finance journals, the inclusion of citations to these articles and journals is beyond the scope of this study.
The data include the total number of citations from the thirteen insurance and actuarial journals and sixteen finance journals during the years 1991 through 1995. Ten of the thirteen insurance and actuarial journals chosen were found in Cox and Gustavson (1990) and Chung and Puelz (1992). The inclusion of the other two actuarial journals was prompted by correspondence with individuals at the Society of Actuaries, and the Journal of Risk and Uncertainty (JRU) was included at the suggestion of several colleagues.
The selection of journals for the study is validated by Baur, Zivney, and Wells (1996), who conduct a comprehensive study on university library subscription policies. All six of the journals listed in their study as "rigorously refereed" and seven of the ten journals listed as "other refereed" are included in the study. Transactions - Society of Actuaries is not included because it only permits members of the Society of Actuaries to publish in it and includes much material that is not in the usual journal article format. One of the other two journals not included in the study is the Journal of Actuarial Practice (JAP), to which only 2 percent of all AACSB-accredited schools subscribe (Baur, Zivney, and Wells, 1996). The finance journals reviewed in the study are the ones reviewed by Zivney and Reichenstein (1994).(6) These journals are those that appear in Heck's (1989) Finance Literature Index.
As stated earlier, the SSCI does not include all of the insurance and actuarial journals relevant to this study and, therefore, the citation data were gathered by reviewing the bibliographies of each of the sample journals for references to the insurance and actuarial journals included in the study. Citations to working papers that were published in one of the insurance or actuarial journals subsequent to the citation were not recorded.(7) Data gathered include the author, journal edition, and page numbers of the cited article as well as the journal edition and page number of the citing article. Only citations from feature articles, short articles, and notes and communications regarding research are included in the data. Citations from comments and replies are not included in the study. Opinion pieces and regular columns like those found in the CLU Journal (CLUJ) and the CPCU Journal (CPCUJ) also were not reviewed for citations.
When conducting a study such as this, it is important to recognize the impact of the time frame from which the sample journals are drawn. Table 1 provides a distribution of the citations recorded by the year in which the cited article was published. By obtaining data from journals published during the years 1991 through 1995 (the most recent five-year period available at the time of data collection), a considerable number of citations recorded are to articles published during the late 1980s and early 1990s. Therefore, by using citation data to evaluate the relative quality of a journal, an unavoidable time lag exists between the time period that is actually evaluated (the late 1980s and early 1990s) and the period of data collection (1991 through 1995). As a result, incremental changes in journal quality that have occurred since the early 1990s are not likely to appear in the results.
INFERENCES DRAWN FROM PRIOR RESEARCH
Borokhovich, Bricker, and Simkins (1994, p. 714) define a journal core as "that set of journals that has the greatest influence on the discipline." Their study of the interjournal communication between finance journals determines that the Journal of Finance and the Journal of Financial Economics comprise the journal core of finance. By analyzing the citation frequency of the insurance and actuarial journals, it is likely that the journal(s) that make up the journal core of insurance and actuarial research can be identified.
As stated above, Baur, Zivney, and Wells (1996), in their analysis of the insurance and actuarial journal subscription practices of colleges and universities, make a distinction between "rigorously refereed" journals and "other refereed" journals.(8) Research that has been subject to rigorous scrutiny by editors and other accomplished researchers is expected to have a more significant impact on future research. Assuming that the classification provided by Baur, Zivney, and Wells is accurate, then it is expected that those journals that are "rigorously refereed" would have higher relative citation rates than the journals classified as "other refereed." Therefore, it is expected that the Astin Bulletin (AB), IME, the GPT, the JRI, the JRU, and the Scandinavian/Actuarial Journal (SAJ) would have relatively high citation rates. This list of "rigorously refereed" journals also is expected to include the journal(s) that make up the journal core.
Baur, Zivney, and Wells find that, for the sample schools that offer a major in insurance, the JRI has a subscription rate that is over two times that of any other "rigorously refereed" journal. If subscription rates are an accurate proxy for citation frequency, then the JRI is expected to be one of the journals comprising the journal core. Provided that the JRI is one of the journals comprising the journal core, the JRI should have a very high citation rate relative to the other sample journals.
Also, given the subscription rates found by Baur, Zivney, and Wells, among the "rigorously refereed" journals, the JRU, the SAJ, and IME are expected to have similar citation rates, and the AB and the GPT are expected to be cited with less relative frequency. Finally, among the "other refereed" journals, the findings of Baur, Zivney, and Wells suggest that the CLUJ and the CPCUJ would have the highest relative citation rates, the Journal of Insurance Regulation (JIR) and the Journal of Insurance Issues (JII) would have a considerably lower citation rate, followed by Benefits Quarterly (BQ), the Journal of the Institute of Actuaries (JIA) and the Geneva Papers on Risk and Insurance: Issues and Practice (GPIP).
Table 1 Distribution of Citations by Year of Article Cited Cumulative Year Percentage Percentage 1995 0.45 0.45 1994 2.41 2.87 1993 5.14 8.01 1992 8.49 16.50 1991 9.49 25.99 1990 9.66 35.65 1989 9.20 44.85 1988 7.73 52.58 1987 6.28 58.86 1986 5.77 64.63 1985 4.23 68.86 1984 3.78 72.64 1983 3.29 75.93 1982 2.36 78.29 1981 3.52 81.81 Pre-1981 18.19 100.00
Borokhovich, Bricker, and Simkins (1994) note that most finance journals tend to cite themselves with a relatively high degree of frequency.(9) They also provide three hypotheses to explain the differences between the self-citation rates across journals. First, they suggest that it is possible that journals encourage self-citations as a form of self-promotion. This hypothesis suggests that journals with a higher rate of self-citations have a higher level of self-promotion. Second, a journal that publishes highly influential articles is likely to be cited with more frequency by outside journals, explaining the differences in self-citation rates among various journals. Finally, a journal that publishes research that is of a fairly specialized nature is likely to have a higher self-citation rate than a journal publishing research that is of a broader nature. This is because a journal publishing in a specialized area is likely to be one of only a few sources from which to reference previous research.
The research interests of authors publishing in each of the sample journals also is likely to affect the results obtained. Journals publishing articles by authors with similar research interests are expected to cite one another with a high rate of frequency. Colquitt, Dumm, and Gustavson (1997), in a study measuring insurance and actuarial research productivity during the years 1987 through 1996, observe the frequency with which authors publishing in any one of the sample insurance and actuarial journals also publish in the other sample journals.(10) They find a relatively high degree of overlap among authors who have had articles published in the AB, IME, and the SAJ.
Colquitt, Dumm, and Gustavson also find a relatively high degree of overlap among authors with articles published in the JRI, the JIR, the JII, the CPCUJ, and the CLUJ. Assuming that the same type of self-promotion that Borokhovich, Bricker, and Simkins hypothesize to exist among journals also exists among authors, then it is expected that journals sharing many of the same authors would tend to cite one another with a high degree of frequency. Interestingly, the JRI is the only journal in the Colquitt, Dumm, and Gustavson study that jointly shares at least one author with every other sample journal. This is further support for the expectation that the JRI is one of the journals comprising the journal core.
Colquitt, Dumm, and Gustavson find very little overlap among authors publishing articles in the JRU and authors publishing articles in the other sample journals. This suggests that the research interests of many of the authors publishing in the JRU do not overlap significantly with the research interests of authors publishing in the other sample journals. Despite the fact that the JRU is listed as a "rigorously refereed" journal by Baur, Zivney, and Wells, the lack of author overlap found by Colquitt, Dumm, and Gustavson supports the expectation that the JRU is not cited with a high degree of frequency by the other sample journals.
Finally, Colquitt, Dumm, and Gustavson find that a very small percentage of authors that publish articles in BQ also publish articles in any of the other sample journals. This lack of author overlap, along with the narrowly defined focus of BQ (which deals primarily with issues related to employee benefits), supports the expectation of a low rate of citation frequency for BQ.
DISCUSSION OF RESULTS
The citations from each of the sample journals were tabulated and the results are reported in Table 2. Table 2 provides a breakdown of the frequency with which each sample journal was cited and where the citations to each journal were found.(11) In addition, the total source articles and the number of references to sources other than the sample journals are provided.
Consistent with Borokhovich, Bricker, and Simkins, we find that the journal most frequently cited by the majority of citing journals is the citing journal itself. Two exceptions to this are the JII and the JIR. These two journals cited the JRI more frequently than any other journal. However, in these two cases, the next most frequently cited journal is the citing journal itself. Also, the JRI is the journal second most frequently cited by BQ, the CLUJ, the CPCUJ, the GPIP, and the JRU. In addition, the JRI is the insurance or actuarial journal most frequently cited by the sample of finance journals, garnering over 73 percent of all the citations to the sample insurance and actuarial journals.
Each journal's self-citation rate is found in Table 2 and each journal's self-citation index is provided in Table 3.(12) The higher the self-citation index, the higher a journal's frequency of self-citations relative to the frequency with which it is cited by the other sample journals.
Among the "rigorously refereed" journals, the JRI (0.40) and the IME (0.67) have the lowest self-citation indices. This suggests that these journals are among the more influential journals in the sample. Among the "rigorously refereed" journals, the JRU (2.16) has the highest self-citation index. Although the high self-citation index could be suggestive of self-promotion by the JRU, it also is consistent with the expectation that the JRU publishes research that is of a specialized nature and unrelated to much of the research published in the other sample journals.
Consistent with the expectation that the "rigorously refereed" journals publish the most influential research, all of the "rigorously refereed" journals, with the exception of the JRU (which is hypothesized to publish research that is of a specialized nature), have a lower self-citation index than any of the "other refereed" journals. Finally, among the "other refereed" journals, the JIR (1.48) and the CPCUJ (1.51) have the lowest self-citation indices, and the CLUJ has the highest self-citation index (4.61).
Table 4 provides a ranking of the journals based upon the total number of citations found among the sample journals, with and without controlling for self-citations. Including self-citations, the JRI (1,076 citations) is the most frequently cited journal with almost 2.5 times the number of citations to the second most frequently cited journal, the AB (434 citations). This is not surprising, given that the JRI is the journal either first or second most frequently cited by eight of the thirteen insurance and actuarial journals and first among the finance journals reviewed. [TABULAR DATA FOR TABLE 2 OMITTED] [TABULAR DATA FOR TABLE 3 OMITTED] [TABULAR DATA FOR TABLE 4 OMITTED] In addition, the JRI is the only insurance or actuarial journal with at least one citation by every other sample insurance or actuarial journal.
The number of citations to journals three through six are very close, with the range between the third ranked journal (IME) and the sixth ranked journal (JIA) being only 73 citations. Once again, the low ranking of BQ is likely attributed, in part, to the specific nature of the journal and the fact that other benefits-related journals are not included in the sample of citing journals.
Excluding self-citations, the JRI and the AB remain the two most frequently cited journals with 598 and 246 citations, respectively. The order of the other ten journals remains fairly consistent with the order when including the self-citations. Only the JRU and the CPCUJ move more than two places. When excluding self-citations, the JRU falls from fourth to ninth and the CPCUJ climbs from ninth to sixth.
Following the Finance Impact Factor devised by Zivney and Reichenstein, an Insurance Impact Factor (IIF) also was calculated for each of the sample insurance and actuarial journals.(13) The IIF measures the average impact of a journal's articles [TABULAR DATA FOR TABLE 5 OMITTED] on the research published in the sample journals. The results of the IIF are found in Table 5.
With regard to the "rigorously refereed" and "other refereed" classifications, the IIF is consistent with the previous results with one exception, the JIA ("other refereed") has a higher IIF than the IME ("rigorously refereed") when calculating the IIF using all citations. However, the JIA falls below the IME when self-citations are not included.
Within the two "refereed" classifications, there are some differences between the rankings based on raw citation counts and the IIF. Among the "rigorously refereed" journals, one noticeable difference is that both the AB and the JRU have a higher IIF than does the JRI. However, when excluding self-citations from the calculation, the JRI moves ahead of the JRU but remains behind the AB. Among the "other refereed" journals, the JII has a higher IIF than the CPCUJ and the CLUJ. This is an improvement for the JII relative to the rankings based on raw citation count.
Table 6 provides a comparison between the findings of this study and those of Baur, Zivney, and Wells, McNamara and Kolbe, and Outreville and Malouin.(14) The JRI is consistently the highest ranking journal in each of the four studies.
Beyond the absolute agreement in the ranking of the JRI, the placement of the journals classified as "rigorously refereed" are not absolutely consistent between studies. All of the "rigorously refereed" journals have the highest citation frequency in this study. In addition, McNamara and Kolbe find that the "rigorously refereed" journals also received the highest ratings of all the sample journals by the deans surveyed.
[TABULAR DATA FOR TABLE 6 OMITTED]
In the Outreville and Malouin study, most of the "rigorously refereed" journals are more highly ranked than the "other refereed" journals. However, Outreville and Malouin find the JIA to be the second highest ranking insurance or actuarial journal among the sample journals. Finally, the subscription rates of the Baur, Zivney, and Wells study are the least consistent with the "refereed" classifications. Although the JRI has the highest subscription rates of the sample journals, the CLUJ and the CPCUJ both have higher subscription rates than any of the other "rigorously refereed" journals. Also, the AB and the GPT, which are "rigorously refereed," have two of the lowest subscription rates of any of the sample journals.
There is strong evidence to suggest that the JRI alone comprises the journal core of insurance research. First, the JRI has almost 2.5 times the number of citations to any of the other sample journals. Second, the JRI is the only journal to receive at least one citation from each of the other sample insurance and actuarial journals. Third, the JRI has the lowest self-citation rate of any of the sample insurance and actuarial journals. Fourth, of the 128 citations to the sample insurance and actuarial journals found in the sample finance journals, over 73 percent of them are to the JRI. Finally, among the four insurance and actuarial journal studies observed, the only consistent finding is that the JRI is the top ranked insurance or actuarial journal among all the journals evaluated.
In addition to knowing the relative significance of journals, there also is value to journal editors, researchers, and teaching faculty in knowing the relative significance of journal articles in a particular field of study. Citation analysis of journal articles is useful to journal editors because it allows the journal editor to see what past articles and subjects in insurance and actuarial science research have had the most significant impact on future research. This information can assist the editor in making future editing decisions on research submitted for publication.
For similar reasons, researchers in insurance and actuarial sciences also benefit from the results of citation analysis of journal articles. With knowledge of the most frequently cited articles, the researcher has information on the types of articles (subjects, methodology, style, etc.) that have been the most influential in insurance and actuarial research. By reviewing some of the most influential research of the past, the researcher likely would be more effective at conducting future research that also is more influential. Finally, graduate programs in business typically involve seminar courses that include, as required reading, those articles that the faculty has determined to be among the most influential articles in a particular field of study. The results of citation analysis of insurance and actuarial journal articles could help professors teaching insurance or actuarial seminar courses to determine what articles would be appropriate to include in graduate instruction.
For each citation to an article published by the sample insurance and actuarial journals, the title of the cited article was recorded. The total number of citations to each of the journal articles was tabulated and the results of the findings are found in Tables 7 through 10.
The most frequently cited JRI articles published in each year, 1980 through 1993, are found in Table 7. The year, author(s), page numbers, title of the article, and the number of citations found in the sample journals are provided. No JRI article published during the years 1994 and 1995 is cited more than twice. For the JRI articles published during the years 1980 through 1993, the JRI article in the table with the fewest citations (five) is the article by Schlesinger and Schulenberg published in 1993.(15)
Cummins and Harrington are the only authors to be credited with multiple articles on the list of the most frequently cited JRI articles for each year. Cummins is a coauthor with Outreville on the most frequently cited article published in 1987 and is the sole author of the most frequently cited articles published in the years 1990 and 1991. Harrington is a coauthor with Nelson on the most frequently cited article published in 1986 and the sole author of the most frequently cited articles published in 1982 and 1984.
The most frequently cited JRI articles, regardless of the year published, are listed in Table 8. The author(s), year/volume, pages, title of the article, and the number of citations are provided. Grabowski, Viscusi, and Evans (1989) and Cummins and Outreville (1987) are cited more frequently by the sample journals than any other JRI article. Eleven of the fifteen articles making the list of the top JRI articles were published during the 1980s. The only three articles from the 1990s to make the list are Cummins (1990), Cummins (1991), and BarNiv and Hershbarger (1990). The only article from the 1970s to make the list is Biger and Kahane (1978).
Cummins leads all authors with four articles on the list of the most frequently cited JRI articles, regardless of the year published. These articles include the articles previously mentioned and found in Table 7 as well as his 1985 article coauthored with Harrington. The only other author with multiple articles on the list is Harrington, with the 1985 article above, an article published in 1984, and an article published in 1986 with Nelson.
Table 9 lists the most frequently cited articles published in each of the insurance and actuarial journals.(16) The journal, author(s), year/pages, title of the article, and the total number of citations are provided. All but one of the most frequently cited articles for each of the journals listed were published during the 1980s or 1990s. It is interesting to note that the article not written in the 1980s or 1990s is Redington's JIA article published in 1952 and is 29 years older than the next oldest article, Panjer's 1981 AB article. Another interesting finding is that 64 percent (7 of 11) of the citations to Merton (1989) and 50 percent (5 of 10) of the citations to Redington (1952) are found in the sample of finance journals.
[TABULAR DATA FOR TABLE 7 OMITTED]
[TABULAR DATA FOR TABLE 8 OMITTED]
[TABULAR DATA FOR TABLE 9 OMITTED]
Table 10 lists the most frequently cited articles published by any of the sample insurance and actuarial journals. Panjer's 1981 AB article, "Recursive Evaluation of a Family of Compound Distributions," is the most frequently cited article with 29 citations. Also making the list are Jewell and Sundt, at a tie for number eight with their 1981 AB article, "Further Results on Recursive Evaluation of Compound Distributions." Interestingly, these two articles were published in the same volume of the AB, with the Jewell and Sundt article appearing immediately after Panjer's article.
Cummins leads all authors with three articles in the list of the most frequently cited articles (JRI/1987, JRI/1985, JRI/1990). Viscusi (JRI/1989, JRU/1989) and Harrington (JRI/1984, JRI/1985) are represented with two articles each. The JRI leads all journals with eight of the top eighteen articles. The remaining ten articles consist of five JRU articles, three AB articles, one IME article, and one GP article.
The bibliographies of articles from thirteen insurance and actuarial journals and sixteen of the leading finance journals during the years 1991 through 1995 were reviewed to establish the rankings of the thirteen insurance and actuarial journals in the sample. Although the number of finance journals reviewed (80 journal years) is larger than the number of insurance and actuarial journals reviewed (65 journal years), the number of citations found in the finance journals make up only 3.75 percent of the total citations to the sample journals.
The insurance and actuarial journals are ranked based upon the journals' total number of citations and their Insurance Impact Factors. The top two insurance or actuarial journals based on the number of citations from the sample journals are the Journal of Risk and Insurance and the Astin Bulletin. The rankings of the journals change very little when controlling for self-citations. Only the positions of the Journal of Risk and Uncertainty and the CPCUJ change considerably (more than two places) when controlling for self-citations. In addition, the JRI is the only insurance or actuarial journal cited by each of the other insurance and actuarial journals reviewed. The journals with the highest insurance impact factor are the AB, JRU, and the JRI. When controlling for self-citations, only the position of the JRU changes considerably (falling four places).
A comparison is made between the rankings generated in this study and those generated by Baur, Zivney, and Wells, McNamara and Kolbe, and Outreville and Malouin. Based upon the "refereed" classifications made by Baur, Zivney, and Wells, the results of this study, McNamara and Kolbe, and Outreville and Malouin are generally consistent with one another. However, "refereed" classifications do not consistently follow the subscription rates of the sample journals. The only absolutely consistent finding of each of the above studies is the number one ranking of the JRI. Based upon the results generated by this and other studies, it is concluded that the JRI is the only journal comprising the journal core of risk and insurance research.
[TABULAR DATA FOR TABLE 10 OMITTED]
Finally, the most frequently cited articles also are reported. The JRI articles most frequently cited by the sample journals are Grabowski, Viscusi, and Evans (1989) and Cummins and Harrington (1987), both with 20 citations. The article most frequently cited by the sample journals is Panjer's AB article published in 1981 with 29 citations. Among the list of the eighteen articles most frequently cited by the sample journals, the JRI leads all journals with eight articles, followed by the JRU with five articles and the AB with three articles.
APPENDIX JOURNALS INCLUDED IN THE STUDY Insurance and Actuarial Journals Astin Bulletin (AB) Benefits Quarterly (BQ) CLU Journal (CLUJ) CPCU Journal (CPCUJ) Geneva Papers on Risk and Insurance: Geneva Papers on Risk and Issues and Practice (GPIP) Insurance Theory (GPT) Insurance: Mathematics and Economics Journal of the Institute of (IME) Actuaries (JIA) Journal of Insurance Issues (JII) Journal of Insurance Regulation (JIR) Journal of Risk and Insurance (JRI) Journal of Risk and Uncertainty (JRU) Scandinavian Actuarial Journal (SAJ) Finance Journals Financial Analysts Journal Financial Management Financial Review Journal of Banking and Finance Journal of Business Journal of Business Finance and Accounting Journal of Finance Journal of Financial and Quantitative Analysis Journal of Financial Economics Journal of Financial Research Journal of Financial Services Journal of Futures Markets Research Journal of International Money and Journal of Money, Credit and Finance Banking Journal of Portfolio Management Review of Financial Studies
The author gratefully acknowledges information provided by Randy Dumm, Sandra Gustavson, Rob Hoyt, Zachary Rolnick of Kluwer Academic Publishers, and Harris Schlesinger. The author also appreciates valuable comments provided by the Editor, two anonymous referees, Randy Dumm, Rob Hoyt, and David Sommer.
1 The subgroups of survey participants were (1) those teaching both insurance and finance in a Ph.D.-granting department, (2) those teaching insurance and finance in a non-Ph.D.-granting department, (3) those teaching only insurance in a Ph.D.-granting department, and (4) those teaching only insurance in a non-Ph.D.-granting department.
2 More specifically, they find that, among their sample of six insurance journals, the number of authors publishing n coauthorship-adjusted articles is 1/[n.sup.2.45] of those publishing one article. For example, given this relationship, the number of authors publishing two coauthorship-adjusted articles is 0.183 times (1/[2.sub.2.45]) the number of authors publishing just one article. With knowledge of this relationship, any given research publication record can be measured against the records of those authors included in the sample.
3 Baur, Zivney, and Wells (1996, p. 485) imply that citation analysis is a preferred method of assessing journal quality but that university library subscription rates serve as a proxy that has been shown to be correlated with citation rankings.
4 The impact factors for the risk and insurance journals indexed by the SSCI in 1995 are: GPT, 0.421; IME, 0.267; JRI, 0.317; and JRU, 2.196.
5 One problem with using the SSCI to influence the significance of insurance and actuarial journals on a narrowly defined area of research (such as insurance, actuarial, and finance research) is that the SSCI bases its rankings on citations from a wide range of journals in the social sciences, not just insurance, actuarial, and finance journals.
6 Although the choice of finance journals to be reviewed may have some impact on the outcome of the rankings of the journals, it is unlikely that a significant change would occur if other finance journals were reviewed. In their study of citations found in the Journal of Finance, the Journal of Financial Economics, the Journal of Financial and Quantitative Analysis, and the Review of Financial Studies from January 1987 through March 1991, Alexander and Mabry (1994) find that no insurance journal made the list of the top 50 journals or other media cited (with at least 22 citations).
7 An exception to this is when an insurance or actuarial journal had accepted an article for publication and it was "forthcoming" in that journal.
8 Baur, Zivney, and Wells classify the Journal of Risk and Insurance, the Journal of Risk and Uncertainty, the Scandinavian Actuarial Journal, Insurance: Mathematics and Economics, the Astin Bulletin, and the Geneva Papers on Risk and Insurance Theory as "rigorously refereed." They classify the CLU Journal, the CPCU Journal, Benefits Quarterly, the Journal of Insurance Regulation, the Journal of Insurance Issues, the Journal of the Institute of Actuaries, and the Geneva Papers on Risk and Insurance: Issues and Practice as "other refereed." From this point forward, this Baur-Zivney-Wells classification will be used without further reference to that study.
9 Zivney and Reichenstein (1994) review the citations from 18 finance journals and also find that the journal most frequently cited by three of the top ten journals is the citing journal itself. Of the remaining seven journals, the citing journal is the second most frequently cited journal for three journals and the third most frequently cited journal for three journals. The frequency with which the Journal of Finance and the Journal of Financial Economics are cited by the sample journals accounts for the majority of instances where a journal does not cite itself with the most frequency.
Alexander and Mabry (1994) review the citations from four of the top finance journals and find that, of the four journals, only one citing journal (Journal of Finance) is the most frequently cited journal. For the other three journals, the Journal of Finance is the most frequently cited journal, and the second most frequently cited journal is the citing journal itself.
10 The sample insurance and actuarial journals that comprise the Colquitt, Dumm, and Gustavson study are the same journals reviewed in this study.
11 The citations credited to the GPT and the GPIP were combined due to the frequency with which these journals were cited as the Geneva Papers (GP), with no distinction made between GPT and GPIP.
12 The calculation of both the self-citation rate and the self-citation index follows that of Borokhovich, Bricker, and Simkins (1994). The self-citation rate is simply the number of self-citations from a journal divided by the total number of citations found in that journal. The self-citation index is the self-citation rate x 100/normalized average citation rate excluding self-citations (per thousand citations).
13 The Insurance Impact Factor equals citations to a journal's articles published in a certain period divided by the number of citable articles published during the same period. The period used for all of the journals except the JRU is 1986 through 1995. The JRU was established in 1987 and, therefore, the period used for the JRU is 1987 through 995. The GP was only one journal until 1990, when the GP was divided into the GPT and the GPIP. The post-1989 citations and articles for the GPT and the GPIP were combined with those from the pre-1990 GP to calculate an impact factor for the GP.
14 The GPIP and the GPT are combined in the current study and in the McNamara and Kolbe study but not in the Baur, Zivney, and Wells study. At the time of the Outreville and Malouin study, the GP had not yet been divided into the GPT and the GPIP. As a result, comparisons between the studies regarding the GP are, in some cases, inappropriate. In addition, not every study included each of the sample insurance and actuarial journals found in the current study.
15 The reason for the lack of citations to articles published in 1993 and later is because no article is credited with a citation until it has been accepted by a journal. Although it is possible that an article published in 1993 might have received citations as a working paper prior to publication, citations to working papers are not included in the results of the study. As a result, an article published in 1993 is not likely to be credited with any citations prior to 1993.
16 Articles from the CLUJ and the CPCUJ are not included in this table due to the incomplete nature of citations provided by the CLUJ and the CPCUJ.
Alexander, John C. Jr. and Rodney H. Mabry, 1994, Relative Significance of Journals, Authors, and Articles Cited in Financial Research, Journal of Finance, 49: 697-712.
Baur, Michael N, Terry L. Zivney, and Grant J. Wells, 1996, The Academic Community's Revealed Preferences Among Insurance Periodicals, Journal of Risk and Insurance, 63: 485-500.
Borokhovich, Kenneth A., Robert J. Bricker, and Betty J. Simkins, 1994, Journal Communication and Influence in Financial Research, Journal of Finance, 49: 713-725.
Borokhovich, Kenneth A., Robert J. Bricker, Kelly R. Brunarski, and Betty J. Simkins, 1995, Finance Research Productivity and Influence, Journal of Finance, 50: 1691-1717.
Chung, Kee H. and Robert Puelz, 1992, An Empirical Regularity in the Market for Risk and Insurance Research Output, Journal of Risk and Insurance, 59: 489-498.
Colquitt, L. Lee, Randy Dumm, and Sandra G. Gustavson, 1997, An Analysis of Risk and Insurance Research Productivity: 1987 Through 1996, Working Paper, Auburn University, Auburn, Alabama.
Cox, Larry A. and Sandra G. Gustavson, 1990, Leading Contributors to Insurance Research, Journal of Risk and Insurance, 57: 260-281.
Gardner, Lisa A. and Joan T. Schmit, 1995, Collegiate Risk Management and Insurance Education, Journal of Risk and Insurance, 62: 625-648.
Heck, Jean Louis, 1989, Finance Literature Index, Second Edition (New York: McGraw-Hill).
McNamara, Michael J. and Phillip T. Kolbe, 1996, The Rankings of Risk Management and Insurance Journals and Tenure and Promotion Requirements at Business Schools, Journal of Insurance Issues, 19:183-191.
Outreville, J. Francois and Jean-Louis Malouin, 1985, What Are the Major Journals that Members of ARIA Read? Journal of Risk and Insurance, 52: 723-733.
Schwert, G. William, 1993, The Journal of Financial Economics: A Retrospective Evaluation (1974-91), Journal of Financial Economics, 33, 369-424.
Zivney, Terry L. and William Reichenstein, 1994, The Pecking Order in Finance Journals, Financial Practice and Education, 4: 77-87.
L. Lee Colquitt is Assistant Professor of Rich and Insurance at Auburn University.
|Printer friendly Cite/link Email Feedback|
|Author:||Colquitt, L. Lee|
|Publication:||Journal of Risk and Insurance|
|Date:||Sep 1, 1997|
|Previous Article:||Valuation of early exercisable interest rate guarantees.|
|Next Article:||Agency theory and life insurer ownership structure.|