Annual Report and Editorial Commentary for The Accounting Review
报告了《会计评论》2013-2014年度的运营情况，包括投稿量、接受率、审稿周期等统计数据，并讨论了编辑政策、稿件积压及作者分布，适合关注会计学术期刊运作的学者和编辑。
This annual report describes the operations of The Accounting Review during the final year (6/1/2013–5/31/2014) of my three-year term as senior editor, with Stacy Hoffman as editorial assistant. The report represents the sixth edition of a reporting format adopted by the AAA in 2009. Along with summary statistics on The Accounting Review (TAR) operations, I will continue the tradition that Steve Kachelmeier started of providing a commentary for our constituents, particularly the AAA Publications Committee and Board of Directors, the new editorial team headed by Mark DeFond, with Elizabeth Garrett as Editorial Assistant, as well as co-editors, authors, reviewers, and readers of TAR and AAA members. Your comments and questions are welcome.Our third year continued the increasing reliance on the online AllenTrack system, which is evolving further under Mark and Elizabeth. For additional details on our operational processes during our regime, please see the 2013 TAR annual report (Evans 2013). That report also discusses the integrity of the academic research process and the integrity of peer-reviewed journals in that process. In my opinion, these issues become more important each year as more reports of fraudulent research in academic journals become public.To facilitate comparisons over time, the 2014 report follows the structure of the 2009–2013 reports. The annual TAR workflow has remained relatively consistent during the 2010–2014 period, after an initial adjustment in Steve Kachelmeier's first year in 2009.Table 1, Panel A reports TAR's comparative workflow statistics for the six years 2009–2014, with the data on a journal-year basis of June 1 through May 31 of the following year. Panel A shows that in 2014 the volume of new submissions increased by 7.3 percent from 543 in 2013 to 583 in 2014. During the first year of the Kachelmeier term, TAR experienced a surge of new submissions, followed by a slight decline in Steve's final two years. Our regime experienced a similar surge in our first year of 2012 relative to 2011, but even further growth in the final year of 2014. The year-over-year growth rates in new submissions versus the corresponding years in the preceding regime have been 2.0 percent, 8.2 percent, and 17.8 percent, respectively. This pattern of increases suggests a generally healthy position of the journal, in terms of this single measure of attracting new submissions. The other columns of Table 1, Panel A show that 2014 achieved new highs in the activity measures of total manuscripts available for evaluation (column (d)), decision letters sent (column (e)) and ending inventory of manuscripts (column (f)).Next, to permit comparability to years before 2009, Table 1, Panel B provides data on a calendar year (CY) basis beginning in 1998. Panel B indicates that the total of 561 new manuscripts submitted to The Accounting Review in CY2013 is exceeded only by the 582 new submissions in CY2011. The third and fourth columns of Table 1, Panel B track how the increased submissions prompted corresponding increases from four TAR issues annually in 2005 to six issues annually starting in 2008, and associated growth in the total annual published pages.Table 2, Panel A reports the editorial outcomes communicated in the total decision letters (column (a)) generated during each of the fiscal or journal years 2009–2014. The final two columns of Panel A of Table 2 use the data in columns (a) to (d) to generate two estimated annual “acceptance rates” for each of the last six fiscal years. For 2014, Acceptance Rate 1 in column (e) divides the 81 acceptances and conditional acceptances in 2014 by the 81 (column (d)) + 466 (column (b)) = 547 “final outcome” decisions, yielding the Acceptance Rate 1 of 14.8 percent. Acceptance Rate 2 in column (f) retains the same 81 acceptances in 2014 in the numerator but now adds to the denominator the 239 “Revise and ‘Uncertain' Decisions” in 2014, which yields an Acceptance Rate 2 of 10.3 percent (81/786). Acceptance Rate 1 can be viewed as an upward-biased measure of the 2014 acceptance rate, whereas Rate 2 is downward-biased, such that the “true” acceptance rate falls somewhere in between—roughly in the 12–13 percent range for 2014. The overall results for 2014 in Table 2, Panel A are largely similar to those for 2011–2013, consistent with the current editorial team's decision to generally retain policies from the preceding regime, together with a generally similar experience in the submission and review process in 2014 as in the previous five years.Table 2, Panel B presents a slightly different “annual cohort” perspective on acceptance rates. Whereas Panel A focuses on the annual flow of manuscripts and editorial decisions in a given year independent of when those manuscripts were initially submitted, Panel B treats each year's set of new submissions as a unique “cohort” and tracks the eventual outcomes for that cohort over the next several years. Thus, the first two lines of Table 2, Panel B shows that for the journal years ending May 31, 2009 and 2010, TAR received 557 and 502 new submissions, respectively (column (a)), all of which, as column (d) shows, were either accepted or rejected as of five years later on May 31, 2014, yielding acceptance rates of 17.4 percent and 13.7 percent, respectively, for these two cohorts of 2009 and 2010 submissions.The final three rows of Table 2, Panel B report the corresponding figures for the status of the 2011, 2012, and 2013 submission cohorts. The final two columns present lower and upper bounds on the annual acceptance rates for these three cohorts of new submissions based on different assumptions concerning how many of the manuscripts that remained in process as of 5/31/2014 will ultimately be accepted. The results show that the final acceptance rate for the 2011 cohort of new submissions will fall in the 14.3 percent to 16.3 percent range versus 15.3 percent to 18.8 percent and 5.2 percent to 19.5 percent for the 2012 and 2013 cohorts, respectively. The wider range for the 2013 cohort reflects the fact that a larger percentage of these manuscripts remained in process as of 5/31/2014.One final observation concerns the relation between TAR acceptance rates, publication rates, and the resulting backlog of accepted but not-yet-published manuscripts. Steve Kachelmeier's regime accepted a sufficient number of articles to build an approximately six-month backlog by the end of his term. Therefore, Stacy and I inherited an inventory of accepted articles to fill the three issues published in our first six months. Throughout the Harry/Stacy term this backlog grew to approximately ten months, which represents five issues. Based on the data in column (d) of Table 2, Panel A, TAR editors have accepted an average of 75.7 articles per year over the last six years. Given that TAR has published 72 articles per year over this period, an additional 3.7 articles per year have been added to the backlog, consistent with the general description above.Having a sufficient backlog ensures a consistent publication rate and a consistent number of articles per issue in contrast to some earlier years in which TAR published relatively “thin” issues comprised of fewer articles, where “thin” issues can hurt a journal's visibility. On the other hand, having too long of a backlog can result in published articles that are less timely, although the online publication process addresses this concern to some degree. In net, the current backlog seems at least sufficient, and my understanding is that Mark and Elizabeth plan to take action to reduce the backlog, a policy that I endorse.Exhibit 1 provides further details on the 786 editorial decisions reported in Table 2, Panel A for the journal year ending May 31, 2014. Exhibit 1 shows that 574 of the 786 decisions (73 percent) were first-round decisions, while 123 (16 percent) were second-round decisions (first revisions) and the remaining 89 (11 percent) were third-round or later. Of the 574 first-round decisions, Panel A of Exhibit 1 shows that rejection was the most common outcome, accounting for approximately 73 percent (227 + 192 = 419 of 574) of the first-round decisions, while the remaining 27 percent of first-round decisions were revisions if we exclude the 2012 Presidential Scholar Address.1 Panel A also shows that of the 419 first-round rejections, we attributed 227 (54 percent) to insufficient contribution, and the other 192 (46 percent) primarily to validity concerns.Next, Exhibit 1, Panel A shows that of the 154 first-round decisions in 2014 that permitted the authors to submit a revised manuscript, 74 were standard “revise and resubmit” decisions, while the other 80 were more qualified “uncertain” decisions. Both “revise and resubmit” and “uncertain” have outcome risk, but the degree of that risk is substantially higher for an “uncertain” decision. Specifically, an “uncertain” letter informs the author that neither the reviewers nor the editor can envision a viable revision path that would address the identified concerns, but that the editor recognizes that the author might be able to construct such a path. Accordingly, such a letter gives the author an option to revise and resubmit, but without explicitly encouraging the author to do so. The intent is to communicate clearly to the author that withdrawing the manuscript might be in the author's best interest if the author's candid assessment is that the concerns raised cannot feasibly be addressed. Experience indicates that almost all recipients of “uncertain” letters choose to revise and resubmit in spite of the cautions, but the rejection rate on “uncertain” revisions is substantially higher than that for standard invitations to revise and resubmit.Moving to the second-round or “first revision” decisions, Exhibit 1, Panel B shows that of the 123 total second-round outcomes, 12 received conditional acceptances and 68 received invitations for further revision, with seven of these in the more qualified “uncertain” category. The remaining 43 (13 + 30) second-round letters were rejections, which are always painful. Nevertheless, a third-round rejection is even worse. This consideration encourages editors to make difficult decisions on manuscripts that appear to have potential but achieved only limited progress in the first revision.By the time a manuscript gets to the third round or beyond, the odds of success increase dramatically. Exhibit 1, Panel C shows that for these manuscripts 68 of the 89 fiscal 2014 decisions were acceptance or conditional acceptance, a rate of 76 percent. Seventeen manuscripts received a further revise and resubmit during the third or later round, and four manuscripts were rejected at this advanced stage of the process. Although we seek to minimize such late-round rejections by making the tough decisions sooner whenever possible, in some cases further rounds appear to be the most appropriate decision despite the risk. Finally, we note that Panel C includes 70 third-round decisions, 18 fourth-round decisions, and one fifth-round decision. Of the 19 fourth- and fifth-round decisions, 18 were conditional acceptances.Panels A–D in Table 3 compare submissions to acceptances by subject area, research method, and the combination of the two. The results offer important insight concerning patterns and trends, including whether submissions in certain subject areas or using certain methods were more or less likely than others to be published in TAR over the fiscal years 2009–2014. These tables are helpful in responding to conjectures that TAR systematically favors or disfavors particular areas or methods of research, where the conjecture could stem simply from comparing the number of publications across topic areas or research methods. Table 3 provides systematic data by relating acceptance rates to corresponding submission rates. Table 3 counts each study only once, even though many studies go through several rounds of revision before eventually being published or rejected. This approach means that the figures in Table 3 will generally differ from those in Tables 1 and 2, which treat each submitted version of a study as a distinct manuscript.The general pattern in Table 3 indicates that The Accounting Review accepts articles at rates that are very similar to the corresponding submission rates, whether by topic, by method, or by topic crossed with method.2 The overall similarity in submission and acceptance rates is consistent with the journal's policy of not emphasizing one area or method over another, but rather seeking to reflect the broad interests of AAA members. In turn, this general similarity between submission and acceptance rates is consistent with our process of selecting editors and reviewers. By choosing reviewers for each submission who are experts in the area of that submission, we seek to subject each submission to a comparable review process.Authors obviously exhibit self-selection preferences in determining the journals to which they direct their submissions. These decisions by authors are one fundamental determinant of ultimate TAR publication rates. For example, although TAR publishes far fewer manuscripts in the accounting systems area or manuscripts employing field study methods as compared to the number of financial archival manuscripts, this difference in publication rates is driven primarily by differences in submission rates rather than by differences in acceptance rates. To see this, compare the two percentages in each cell of Table 3, Panel D. The first (second) percentage indicates that cell's percentage of all submissions (acceptances).In this way, Table 3, Panel D compares the percentage of all submissions and acceptances by area and method over the last six years, 2009–2014. For example, the two percentages in the cell in Table 3, Panel D for the combination of the financial accounting area and the archival research method are “43.5%” and “(37.5%),” indicating that over the last six years financial archival studies have comprised 43.5 percent of all TAR submissions and 37.5 percent of all TAR acceptances. The resulting [Acceptance Rate − Submission Rate] differential is −6.0 percent, which is the least favorable differential of any cell in Panel D. This pattern is consistent with the observation above that financial archival was the least-favored category by this measure over the 2009–2014 period.In contrast to the preceding example, the overall pattern of submission and acceptance rates in Table 3, Panel D shows a generally close correspondence between the submission and acceptance percentages. Other than the −6.0 percent difference noted above, none of the individual cells (as opposed to “Total” cells in the bottom row and the far right column) in Table 3, Panel D have [Acceptance Rate − Submission Rate] differentials with absolute values greater than 2.1 percent. The only other distinctive pattern in Table 3, Panel D is that accounting studies have greater acceptance rates than submission rates across all four research method resulting in the bottom “Total” row in Panel D that studies for all research methods percent of TAR submissions, but percent of acceptances over the last six years. In my the rate differentials are and in the between submission and acceptance rates of −6.0 percent for financial archival studies and percent for all accounting studies over a could more than preceding of submissions and acceptances across areas is to the general TAR policy of with to a of research areas and methods of the important that the TAR senior editor can with to to research in particular areas or using particular methods is through the of the team of TAR For example, I continued the approach of TAR's to research in the accounting systems area and to studies using methods by and respectively, to as TAR editors to submissions in accounting systems and Table 3, Panel A shows one acceptance in the systems area in 2014, and the same is in Table 3, Panel B with one field and study acceptance in the process of an team of TAR editors is an The from to the to have an editorial team that has subject and method across a range of areas and while at the same time the total number of editors and the across all authors would that one or more of the TAR have and for the author's research area and method, and further would that one of these editors the author's To increase our odds of this I an additional to the total to co-editors, while the of Specifically, I the of editors slightly from the area and the financial archival area of the experience of having to that editors in the financial archival area more than the of new manuscripts per year. Based on our three years of I that these were decisions. of our has very during these three years, and we have generally been able to all submissions while the of new manuscripts including revisions) per per in to The has been in the area, where has an average of more than new manuscripts annually over our three-year a final under our current regime, the senior editor as the for new manuscripts. This means that when we received more manuscripts than we to in a particular area in a given I generally as editor on these manuscripts, which or not be my area of In my this is a result of in the of manuscripts, and reflects that our can In such the senior editor will almost always more on the reviewers than might if the manuscript were in his or area of terms of the volume of decision individual between and decision letters in journal year 2013 1, 2012 to May 31, and between and letters in journal year 2014. The increase reflects that in 2014, the third year of our term, individual an increasing number of decision letters on as well as letters on new submissions, relative to our year of The percentage of all letters that I from approximately percent in journal year 2013 to percent in journal year Panel A the of all authors who have published in The Accounting with columns for 89 year and for years in annual Panel A for by of an to each of an author two Panel A of that author's to Panel A not for such as The of at and of at with larger have an in the A of Table in based on the as of the of TAR Both columns exclude from AAA Presidential Scholar which we by as well as six from research two in 2009, one in 2012, one in and two in 2014. I also note that the TAR policy a TAR or senior editor from as editor on a articles in the column of Table Panel A reflect different with the by from a of articles to a of of an with two for one of the In terms of relative of the different over six years are consistent with a relatively limited in TAR over this The ten in the Table Panel A column for percent of the total articles, which is well the percent for The Accounting Review by and in their of in articles published across four accounting journals and across ten journals from Table reports that TAR and Accounting are less than the other two accounting journals they and that TAR is in when compared ten these results are consistent with only a of in TAR in years.Table Panel B the of TAR authors published in calendar year 2014 and from of each to each of and AAA Presidential Scholar and six Panel B in of the of by The general that I take from Table A and B is that TAR has been relatively in terms of author and author over the years Panel C reports data on for submissions and acceptances from authors with versus remained relatively from fiscal 2013 to fiscal 2014. This is in contrast to the from fiscal 2012 to fiscal which an increasing percentage of submissions and acceptances from the For fiscal 2014, authors the for percent of TAR's submissions, slightly from percent for fiscal and for percent of acceptances versus percent for fiscal Panel D by the fiscal 2014 submissions and acceptances by Of the submissions, percent of total fiscal 2014 unique and percent) the most the an acceptance percentage above their corresponding submission while for other the acceptance percentage is less than or to the submission In terms of for is encouraging that submissions and acceptances now for approximately to of all TAR reports time data for the 786 editorial decisions from June 1, 2013 through May 31, 2014, as the number of from the submission to the when Stacy Hoffman sent the decision Our and for 2014 are and which is very consistent with the of and of in area of in 2014 was to reduce the in of I this issue with the AAA Publications Committee and with the who the TAR during the AAA in Stacy and I followed with several to editors on how we might reduce the at least progress in this in 2014 by the previous which the percentage of increase from percent in 2012 to percent in In 2014, we this percentage to percent. Finally, I note that of the manuscripts with the time, were by three which is consistent with a combination of and as an important of long and Stacy TAR editorial has the on time with a unique combination of and In the that we could have without would have been a less I our and for their as TAR research have identified potential reviewers who are and have in a I very to be able to the who have their for The Accounting Review over this final year of our three-year and I from each of their letters that they have an of time and to to an and appropriate decision and to communicate their decision in a to all manuscript fourth go to all TAR Editorial Board largely the of TAR as a In the same way, many to the additional reviewers in who many of of without financial and one or more for TAR during this I many to Steve Kachelmeier and editorial in the previous regime, for their in Stacy and I Mark and Elizabeth the new editorial who have been a to with as we operations to I the Accounting particularly the many on the AAA in as well as the academic who on the AAA Publications Committee and Board of I particularly AAA and our of AAA for and in with issues to the integrity of the academic research process during these last two years. Stacy and I to our at Elizabeth Garrett TAR editorial and in and for their and their and helpful to our many I the of and the of of and for their to providing Stacy and with the and we I my accounting academic at the for this and of as TAR senior and most I my for their particularly my of years, for being the and that all