The number of papers submitted

In describing my goals to the Executive Council and various other groups a year ago, I expressed both the hope and the expectation that we would immediately begin to receive more manuscripts than had previously been submitted to the APSR. One reason I hoped for such growth was that it could bring in more potentially publishable manuscripts; even if the proportion of promising submissions remained constant, having more of them to consider would itself be a good thing—especially in light of the expanded number of pages that will soon be opening up for articles. No less importantly, I hoped that an enhanced flow of manuscripts would bring in more papers from parts of the discipline that have been underrepresented in the APSR's pages. My expectation that we would in fact begin to receive more submissions was based on my determination to make it known far and wide that we are eager to receive more papers than the APSR has gotten in the past, that we are working to ensure that the review process runs as it should, and that we aspire to publish the best article-length work across the full range of our substantively, theoretically, and methodologically diverse discipline.

This expectation was more than amply fulfilled during 2001–2002, as the data in Table 1 indicate. Under the three editors who preceded me, the number of papers submitted rose annually, from a mean of 426 under Samuel Patterson (1986–1987 through 1991–1992) to 485 under G. Bingham Powell (1992–1993 through 1994–1995)—an increase of 14%—and then to 506 under Ada Finifter (1995–1996 through 2000– 2001)—an increase of 4%.¹

The figures just given tell us something well worth knowing: that 2001– 2002 was a record-setting year insofar as the total number of submissions was concerned. However, these figures are somewhat misleading, for they intermingle resubmissions with original submissions; thus, a single paper that was submitted n times is counted as n papers. For purposes of charting the popularity of the APSR as a potential outlet for authors seeking a good home for their papers, it makes more sense to focus on the number of “original” or underlying papers submitted rather than on the total number of submissions. These figures, shown in the rightmost column of Table 1, indicate even more substantial growth during 2001–2002, to 509 in 2001–2002 from 327 in 2000–2001. That is a 56% one-year increase, or, from Finifter's six-year mean of 381, a 31% increase. Clearly, then, the APSR attracted much greater interest from contributors during 2001–2002 than it had in recent years.

The mix of papers submitted

In light of charges that the APSR ill serves those whose interests fall outside of American politics and that it is overtly or covertly hostile to research not based on formal models or complex statistical analyses, the substantive and methodological mix of the papers we receive and publish are matters of special concern. A major goal of my editorship is to achieve greater diversity in the manuscripts we receive and the articles we publish.

It would be unrealistic to expect immediate wholesale changes in the mix of papers we receive. Suspicions built up over the course of many years are unlikely to be overcome simply by the brave words and good intentions of a new Editor, and even a sudden surge of submissions from a previously underrepresented field (e.g., comparative politics) could be offset by the continuing or enhanced flow of submissions from another field (e.g., American politics).

Categorized in terms of primary analytical approach, the papers we received after our office opened last September closely resembled those submitted to the APSR during Ada Finifter's two terms as Editor. As the top portion of Table 2 indicates, almost half of the papers that we received between September 1, 2001, and August 15, 2002, relied principally on quantitative methods, and another quarter made use of formal modeling, either exclusively or in combination with quantitative analysis. About three papers in ten were classified as interpretive/conceptual; of these, about 60% were normative theory papers. “Small- N” or comparative case study-based submissions were rare.

Some early signs of change can, however, be gleaned from the field-by-field composition of submitted papers, as shown in the bottom portion of Table 2. Before discussing these figures, let me urge caution in interpreting them, given the obvious difficulty of establishing which papers belong in a given field. For example, comparative politics and international relations increasingly blend into one another, and American politics meets international relations in research on American foreign policy. Similarly, even the most arcane formal theory papers are about something, but the “formal theory” category intermixes a methodological criterion with several substantive ones.²

Even with these caveats in mind, there was perceptible movement during 2001–2002 toward a lessening of the dominance of submissions about American politics. This does not mean that the number of American politics-focused papers declined. Because the total number of submitted manuscripts increased sharply, the absolute number of American politics papers increased even though their proportion of the total was dropping. Nor should it be assumed that the American politics submissions were invariably quantitative or formal in approach, for some promising qualitative papers on American politics were also received (a good case in point being Howard Gillman's “How Political Parties Can Use the Courts to Advance Their Agendas: Federal Courts in the United States, 1875–1891,” which appeared in our September 2002 issue). Whether the small increments in the comparative politics and international relations categories represent first steps toward greater diversification remains to be seen. However, the small up-tick in methods-oriented submissions should not be interpreted as a counter-trend, for some of these papers were qualitative rather than quantitative (e.g., Tim Büthe's “Taking Temporality Seriously: Modeling History and the Use of Narratives as Evidence,” also in our September 2002 issue).

Processing

If the trends considered so far amount to a combination of major change (in the number of submissions received), some signs of change (in the distribution of submissions across fields), and little or no change (in the distribution of submissions across analytical approaches), the news concerning manuscript processing is unambiguous.

In line with the goals of simultaneously enhancing the quality of reviews, speeding up the review process, and making it more transparent, we introduced several new review procedures during 2001–2002. For one thing, we began inviting authors to suggest the names of appropriate reviewers of their papers. For another, we began sending anonymous copies of my decision letter and of all the reviews to the reviewers. Moreover, even though we send every paper out to three reviewers, if the first two reviews point clearly toward rejection, we continued my predecessors' practice of rejecting the paper at that point rather than waiting for the third review to arrive. These procedures have been very favorably received by authors and reviewers alike. It is worth noting, however, that although some of these procedures (most notably, declining papers on the basis of two negative reviews) speed up the review process, others (especially sending the decision letters and reviews to the reviewers, which involves a good deal of extra paper-handling) undoubtedly slow things down somewhat.

The data shown in Table 3 indicate that, notwithstanding the start-up status of our office, the disruptions we experienced during the fall, and the unprecedented, large number of submissions we received, our efforts to speed up the review process met with success. We generally managed to get submitted manuscripts out for review within a day of receiving them; our reviewers typically met or bettered the short target dates we requested for completing their reports; and as soon as we had in hand the information I needed to make a decision, I made it. As a consequence, our median turn-around time was just 39 workdays from the day a paper arrived through the day I signed the letter informing the corresponding author of the outcome. If we manage to avoid the disruptions that plagued us in the fall of 2001 and as more and more reviewers submit their reports electronically, we may be able to speed up the process even more. However, this is not a high priority, and given the pace at which the process is already moving, we may be close to the limit of the possible. Indeed, we will probably do well to hold our own in this respect in coming years, especially if the high volume of submissions continues.

To give credit where it is due, I want to emphasize that our ability to process manuscripts in a timely and efficient manner has been greatly enhanced by the time, effort, and care that Ada Finifter and programmer Paul Wolberg devoted to designing and implementing a database program for overseeing the APSR's operations. By graciously making this program available to us and by going out of her way to respond to our numerous requests for her counsel, Ada Finifter not only helped ensure a successful editorial transition but also made a contribution to the APSR that has continued beyond her term as Editor.

Outcomes

In recent years, the APSR has been criticized for overusing “revise and resubmit” invitations—for extending these invitations to too many papers, for too often piling one “revise and resubmit” decision on top of another, and for rejecting too many of the papers that authors had taken the time and trouble to revise, often multiple times.

I am following a conservative course vis-à-vis “revise and resubmit” invitations. In deciding whether to invite a resubmission, I ask whether (most of) the reviewers are genuinely enthusiastic about the paper, and whether the suggested revisions are narrow enough to be readily accommodated within the current framework of the paper. If I can answer “Yes” to both questions, I issue a “revise and resubmit” invitation. If I cannot, I reject the paper.

Following this conservative course during 2001–2002 produced a considerably lower first-round “revise and resubmit” rate than that of preceding years, and a correspondingly higher rejection rate (see Table 4). Of every 100 firstround decisions I made, 85 were rejections and another four (shown as “other” in the table) were decisions not to open the review process in the first place. (The latter decisions were for papers that were too long, not anonymous, or in some other way inappropriate for APSR review. Most of the authors of these papers subsequently submitted corrected versions.) Only 11 of every 100 of my first-round decision letters conveyed an invitation to prepare a revision —almost exactly one-half the rate of first-round “revise-and-resubmit” invitations issued between 1995 and 2000.

According to the second row of Table 4, 25% of my second-round decisions were rejections. That does not, however, mean that I turned down one out of every four of the resubmitted papers that had been issued a “revise and resubmit” in the first round. In fact, every second-round rejection, with only one exception, was of a paper that had received an “other” decision in the first round, had been “fixed” by the author and resubmitted, and had then been sent out in the second round for its initial review. Thus, I rejected only one paper that had been resubmitted after receiving a “revise and resubmit” invitation.³

In most instances, my second-round acceptances were conditional rather than final. In accepting a paper conditionally, I advise the author (1) that I will assess the next version myself rather than sending it out for further review, and (2) that I will accept it as soon as I receive a version of it that settles a few remaining issues. Most often, the point of the remaining revisions is to improve the accessibility and appeal of a paper by reworking its introductory and concluding sections and/or editing the text to enhance readability; on some occasions, relatively minor substantive matters also require attention.

Of obvious interest is the question of whether certain types of papers fare better than others. Table 5 addresses this issue by comparing first-round outcomes; the pattern is virtually identical, but more complicated to present, when decisions from subsequent rounds are considered as well. The answer is that during 2001–2002, rejection rates varied very little across papers representing different approaches to and fields of political science. Certainly rejection was no less likely for papers representing approaches and fields often assumed to enjoy “most favored” status at the APSR (formal and quantitative papers and those focused on American politics) than it was for other papers. Indeed, if anything, papers from supposedly favored approaches and fields were more likely to be rejected, though it would be unwise to make much of these minor differences. The only substantial deviation from the overall pattern was the unusually low first-round rejection rate for the few small-N studies that we received, though the number of submissions on which this percentage is based is extremely small. In any event, if the small-N papers on which I invited resubmissions are ultimately accepted, I hope their publication in the APSR will encourage a considerably larger volume of submissions of this type in the future.

A different way to assess outcomes is simply to see how many papers representing each approach and field were accepted for publication during 2001–2002. For this purpose, Table 6 focuses on papers on which I made the first-round decision, disregarding papers I inherited at a later stage of the review process from my predecessor. Arrayed across fields, the 18 “new” papers I accepted closely followed the field distribution of the papers we received, as a glance back at Table 2 will establish. That is, the more papers submitted in a given field, the more accepted. A somewhat different pattern, however, held in terms of the distribution of accepted papers across analytical approaches. Of the 18 “new” papers I accepted, ten were quantitative or formal and eight were interpretive/conceptual. By contrast, roughly 70% of the papers that were submitted during 2001–2002 were quantitative or formal and just 30% were interpretive/conceptual. Thus, relative to the papers we received during the year, the papers that were accepted actually overrepresented interpretive/conceptual submissions and underrepresented quantitative and formal ones.