ADVANTAGES OF THE SHIFT FROM BOOK REFERENCES TO JOURNAL ARTICLE REFERENCES

Many of the most significant advantages of an increasing reliance on journal articles over books in political science research result from the much shorter publication timeline (Savage and Olejniczak Reference Savage and Olejniczak2021) and the resulting ability of scholars to address contemporary political issues in a timelier manner. For example, manuscripts submitted to American Political Science Review (APSR) can be reviewed and published in as little as three months on FirstView (Tripp and Dion Reference Tripp and Dion2024), whereas books usually take 1.5 to 3 years to publish (Boboc Reference Boboc2025; Saiya Reference Saiya2022). This expedited review and publishing process has the potential to enhance the discipline’s ability to achieve its fundamental goals: creating “reciprocity between the ‘closet philosopher’ and ‘those engaged in the active walks of political life’ and exerting influence on the ‘world of action’” (Goodnow Reference Goodnow1904, 35, quoted in Gunnell Reference Gunnell2006). Stated simply, a greater reliance on journal articles increases the ability of scholars to address and influence current debates and policy decisions.

Moreover, the greater reliance on journal articles can be advantageous for early-career researchers who need to establish their credentials and contribute to ongoing scholarly conversations. Publication metrics, such as impact factors and citation counts, are used increasingly for hiring, tenure, promotion, and funding decisions (Carpenter, Cone, and Sarli Reference Carpenter, Cone and Sarli2014). The ability to publish quickly—and for others to read and cite this work—may help researchers build their reputation, increase their ability to influence current debates and policy decisions, and make a persuasive case about their contributions to the discipline to administrators and rank and tenure committees.

DISADVANTAGES OF THE SHIFT FROM BOOK REFERENCES TO JOURNAL ARTICLE REFERENCES

The previous discussion highlights several potential positive outcomes of an increasing reliance on journal articles over books. However, there also are significant concerns. First, books often offer more comprehensive and in-depth explorations of topics, covering various perspectives, methodologies, and historical contexts. Journal articles, in contrast, are generally more concise and focus on specific research questions or narrow aspects of a field. They frequently prioritize quantitative methodological approaches, sometimes at the expense of the descriptive, qualitative, and historical analyses necessary for a deep understanding of a problem and the formulation of a comprehensive theory. Journal articles also often lack the interdisciplinary insights provided by books. Consequently, a growing reliance on journal articles may create “echo chambers” and stifle the critical thinking essential for broader synthesis and integration of knowledge.

Second, the pressure to publish frequently can lead to rushed or less thoroughly vetted research. The growing number of journals and, in many cases, increases in the number of manuscript submissions have resulted in reviewer fatigue, making it difficult to find qualified reviewers or resulting in less-than-best reviews (Breunig et al. 2015). Although slower to publish, academic books typically undergo rigorous editorial and review processes, contributing to high quality and thoroughly researched content. Ultimately, the greater reliance on journal articles may result in the referencing and propagation of poorly reviewed or bad research.

HYPOTHESES

Recognizing the evolution of academia, including new research questions, changing curricula and other standards, shrinking budgets, the expansion of online journal databases, and the increase in faculty members working remotely, we posit several hypotheses. First, we suggest that scholars have become increasingly reliant on academic journals as an authority for their research. Specifically, we hypothesize that:

The corollary likely follows:

We expect that these changes have been particularly clear since the COVID-19 pandemic (i.e., 2020 and beyond). As entire physical campuses transitioned to online operations—some for multiple semesters—students and faculty members faced prolonged periods without easy access to physical books in libraries, and they adjusted by sourcing reference materials online. At the same time, libraries were asked to allocate additional resources to online journal catalogs, necessitating a reduction in their book budgets (Frederick and Wolff-Eisenberg Reference Frederick and Wolff-Eisenberg2020; Mintz Reference Mintz2024). Furthermore, AI and other tools for accessing and analyzing research and data have grown exponentially. Thus, we posit:

The discipline varies significantly across subfields, each with its own methodologies and research styles. Given this diversity, we anticipate that political theory, which heavily relies on classic texts, will reference a lesser proportion of journal articles (and a greater proportion of books) than other subfields:

The consequences of referencing more journal articles also are important to consider. If scholars increasingly reference journal articles, we should observe an increase in the number of references per article. Although changing patterns of referencing may not be the only factor leading to growth in references—other potential explanations include increasing word counts, exempting references from word counts for publication, and an emphasis on citing authors from marginalized backgrounds—it is likely a contributing factor. The logic behind this hypothesis is straightforward: whereas a book has the space to address A, B, C, D, and E, a journal article may have the space to address only A and B. To fill these voids, scholars must reference additional journal articles to address C, D, and E. Thus, we suggest:

DATA

To examine changes in journal referencing over time, we analyzed data from 1990 through 2024. This timeframe is recent enough that it reflects the modern discipline—works that form the foundation of much of the literature taught in classrooms and cited today. However, it allowed us to consider equally three potentially important phases of citation. First, it includes approximately 10 to 15 years of research published before the widespread availability of online journals and Internet access in faculty offices and homes. For example, two of the most widely used online databases of scholarly articles, JSTOR and ProQuest, launched in 1995 (“Mission and History” 2004; “ProQuest” 2008); libraries did not begin to adopt these resources widely until the latter part of the decade (Watts Reference Watts2024). Second, there was a 10- to 15-year intermediate period, in which digital tools were well established but many faculty members continued to rely heavily on physical library resources. Third, the last five to 10 years of our analysis reflect shifting library budgets and the post-COVID-19 university. There is no doubt that the next wave is approaching rapidly and will reflect additional changes because of AI research tools.

Our study focuses on top-ranked political science journals (Garand and Giles Reference Garand and Giles2003), supplemented by a purposeful sample to explore variations across different subfields and methodologies. Garand and Giles’ rankings were based on a subjective survey of 565 political scientists who were asked to rank the impact, evaluation, and familiarity with more than 100 journals in the discipline. This survey, conducted in the middle of our study period, ranked APSR, American Journal of Political Science, Journal of Politics, World Politics, International Organization, and British Journal of Political Science as the top overall journals in the discipline. To these, we added Perspectives on Politics, an APSA journal established in 2003, and subfield-specific journals (i.e., Comparative Political Studies, International Studies Quarterly, Journal of Conflict Resolution, Journal of Policy Analysis & Management, Journal of Urban Affairs, Public Administration Review, and Political Behavior) based on their rankings in the survey (Garand and Giles Reference Garand and Giles2003, 296).Footnote ¹ Footnote ²

We coded the subfield and the references (manually and using machine learning) of the first 10 articles published by each journal at the beginning and the midpoint of each decade within our study period, as well as for 2024 (i.e., 1990, 1995, 2000, 2005, 2010, 2015, 2020, and 2024).Footnote ³ In total, our analysis included 872 articles with 50,453 citations (Yanus and Ardoin Reference Yanus and Ardoin2025).Footnote ⁴

ANALYSIS

We begin by testing our specific propositions, starting with Hypotheses 1a and 1b. Specifically, we anticipated that during the 34-year study period, references to academic journals became more frequent and references to books decreased.

These changes are illustrated in figure 1 and the full data are presented in table 1. In 1990, books constituted more than half (52.3%) of all references and journals constituted less than 40%. Over time, references to books have decreased steadily, replaced by a growing reliance on journal articles. In 2005, the first year in the collected data that journal references surpassed book references, the proportion of references to each was almost equal (i.e., 48.4% of references were to journals and 45.7% were to books). However, in the subsequent 20 years, journal references have dramatically surpassed book references: in 2024, 65% of references were to journal articles and only 28.4% were to books.Footnote ⁵ A two-sample t-test reveals that the difference between the proportion of book and journal article references in 1990 and 2004 is statistically significant (t= –1025.19, p<0.000).

Table 1

References by Type, 1990–2024

Figure 1

References to Books and Journal Articles, 1990–2024

Hypothesis 1c further suggests that there should be evidence of an additional reliance on journal articles after the 2020 COVID-19 pandemic. This hypothesis also is supported by the data. The percentage of references to journal articles increases by 5 percentage points between 2020 and 2024 (from 60.4% to 65.0%); this difference is both statistically and substantively significant (two-sample t-test; t=-231.24.19, p<0.000). As figure 1 illustrates, this four-year period represents a noteworthy deviation from the relative stability in reference patterns observed between 2010 and 2020. That is, it resembles the period between 2000 and 2010 when online repositories began to be more widely adopted by a wide range of academic libraries and accessed by students and faculty.Footnote ⁶

Variations by Subfield

We did not expect that scholars’ reliance on books and articles would be consistent across subfields. Instead, Hypothesis 2 posits that political theory, which relies more on classic texts, will reference fewer journal articles. The data shown in figure 2 and table 2 provide support for this claim. Only modest differences in the number of references to journal articles by subfield existed before the late-twentieth-century revolution in digital access. However, between 2000 and the 2020 COVID-19 pandemic, political theory articles averaged only about 30% of references from journal articles. The other subfields experienced an upward trajectory in journal references, increasing from about 50% of all references in 2000 to approximately 60% of references in 2020, which is a 20- to 30-percentage-point difference between political theory and the other subfields.

Table 2

References to Journal Articles by Subfield, 1990–2024

Figure 2

References to Journal Articles by Subfield, 1990–2024

During and after the COVID-19 pandemic, however, political theory looks more like other subfields, particularly American and comparative politics. In fact, much of the post-COVID-19 increase in journal references observed in testing Hypothesis 1c is attributable to dramatic changes in political theory and international relations. The proportion of references to journal articles in American and comparative politics and public administration remained almost constant with their 2000 levels, but references to journal articles in political theory and international relations increased by approximately 25%. If this trend persists, it may enhance the subfield’s ability to make timely connections to “those engaged in the active walks of political life” and influence the “world of action” (Goodnow Reference Goodnow1904, 35).

It also is noteworthy that although we do not discuss it in this article, the inverse can be said about book citations. Between 2000 and 2020, approximately 60% of references in political theory articles were to books, whereas book references in other subfields declined from approximately 50% of all references in 2000 to closer to 25% by 2024.

Consequences: Contributing to an Increase in References?

The shift from referencing books to journal articles has numerous consequences for the discipline. Exploring all of these implications is beyond the scope of this study; however, we can examine one significant trend: how an increasing reliance on journal article references affects overall reference counts. Hypothesis 3 suggests that although it is not the only potential cause, a growing reliance on journal articles may contribute to an increase in the average number of references per article.

Table 3 provides clear support for this proposition. In no case did a journal reference fewer sources in 2020 or 2024 than in 1990. A review of the individual-level data finds that only 4% of articles in 1990 included more than 100 references compared to more than 10% in 2024. Overall, most journals increased their overall number of references by about a quarter or a third. However, the increases in some journals—particularly subfield-specific outlets—were much greater. For example, the 1990 articles from the Journal of Conflict Resolution cited only 315 articles; by 2020, this number had increased almost threefold to a total of 878.

Table 3

Total Number of References by Journal, 1990–2024

Figure 3 provides further evidence of these changes. In 1990, the average number of references per article in all of the journals we studied was 43; by 2024, the average number increased to 71, which represents a 65% increase over time. Figure 3 also illustrates the changes in the average number of references per article in the three major journals (i.e., APSR, American Journal of Political Science (AJPS), and Journal of Politics (JOP)), revealing a significant increase for each journal. Potentially because of its higher maximum word count, APSR represents somewhat of an outlier. Even with an average rate of 60.2 references per article in 1990, APSR increased to 84.8 references per article in 2024. It is interesting that three of the 10 APSR articles examined in our 2024 analysis included more than 100 references. Future analyses should explore in depth the potential causes of this increase considering not only the type of references but also the ratio of references to length and the authors whose work is being cited.

Figure 3

Average Number of References by Year and Journal, 1990–2024

FUTURE DIRECTIONS

These preliminary findings document important changes that merit further investigation related to references as well as the scholars who are conducting the research. Regarding how patterns of referencing have changed: (1) exploring in more depth the differences not only in these top-ranked journals but also other niche-specific and lower-ranked journals could shed light on changes in the discipline, training, and institutional resources; and (2) exploring biases (e.g., in authors, publishers, and venues) in the specific books and articles being cited may speak not only to how knowledge is shaped but also to potential effects on tenure and promotion decisions.

With respect to authors, there are many variables to consider about the identity of scholars and the sources that they regard as authorities and/or have the resources to access. For example, are women-identifying faculty members—who typically have greater familial pressures and are more likely to be in contingent and teaching-centered positions—more likely to reference articles than books? Or are biases more institutional than based on descriptive characteristics? For example, do scholars at non-research universities exhibit different citation patterns than those at research institutions? Similarly, how does online, English-language access to journals affect the ability of scholars located outside of the United States to publish in major English-language political science journals? How, if at all, do these trends intersect with the increase in international editorial teams, including at the top three journals (i.e., APSR, AJPS, and JOP) in the discipline? There also are important questions about author teams, including how the number and seniority of coauthors may affect the type and variety of works cited.

Finally, the findings of this study demonstrate how citations have changed but cannot fully explain why. We suspect that availability, access, and financial shifts may be major driving factors behind the growing reliance on journal articles. However, other potential explanations merit investigation. Perhaps shifting faculty workloads and the growing demands of the tenure process (where it is available) encourage scholars to publish their best ideas in journal articles rather than books. Alternately, the growing emphasis of graduate institutions on journal articles over books in their curriculum, including the emergence of the three-article dissertation, may be an important contributing factor. These changes also may be the result of asking and answering new types of questions in the discipline—particularly in some subfields as suggested by the changes in international relations and political theory discussed in this article. Or it may be due to the types of data that researchers collect and analyze, including the increase in large off-the-shelf databases and the lower cost of original survey research. Understanding how these factors affect scholars’ decisions regarding authoritative sources (and, therefore, references) would be an important contribution to guiding how departments, universities, and associations support scholarly work and how, as teachers, we design our courses and curricula moving forward.