OUTLINE OF THE ANALYSIS

To evaluate trends in the topics discussed in political science research, I selected eight leading journals based on US political scientists’ evaluation of political science journals in 2007 (McLean et al. Reference McLean, Blais, Garand and Giles2009). In decreasing order of relative impact, the top-rated journals were American Political Science Review (APSR), American Journal of Political Science (AJPS), Journal of Politics (JOP), British Journal of Political Science (BJPS), International Organization (IO), World Politics (WP), Comparative Political Studies (CPS), and Comparative Politics (CP). ^{Footnote 3} Of these eight, the latter three are considered journals specific to comparative politics (Munck and Snyder Reference Munck and Snyder2007). My decision to search within these top eight journals—which include four general-interest journals and one largely devoted to international affairs—stems from the fact that, in practice, the division of political science research into one of three domains is not hard and fast. ^{Footnote 4} The topics that may be of interest to comparativists overlap into other areas and researchers frequently draw on work from the other subdisciplines. Moreover, ignoring the research published in the top discipline-wide journals would bias the analysis against some of the most prestigious research on topics relevant to comparative politics research. Thus, whereas the approach presented in this article is suitable for examining trends in the discipline as a whole, its focus frames the discussion of topics in the context of comparative politics research. This focus is justified, in part, by the occurrence of three journals among the top eight that are devoted primarily to comparative politics research, whereas only one may be considered exclusive to international relations and none specializes in American politics research. Although it may be beneficial, it is beyond the scope of this article to elucidate on various research agendas across the entire discipline.

For each journal, I downloaded citation information for all articles published from the date of the journal’s founding until 2015. I omitted from the sample any citation for which there was neither author nor title. This eliminated front and back matter, volume information, and letters from the editor, as well as the voluminous notices and messages published during the early decades of APSR. This omission nevertheless retained errata and book reviews, culminating in a sample that contained 25,845 citations and represented 521 published issues between 1906 and 2015.

Figure 1 lists the number of titles and abstracts available by year for each of the eight journals. The first journal to emerge was APSR, which was first published in 1906. For the first few decades of APSR’s existence, the journal published numerous reports and briefings related to political matters that ran the gamut from recent court decisions to trade disputes. With the outbreak of World War II, additional political science journals began to emerge. The first issue of JOP was published in 1939, followed by IO and WP in 1947 and 1948, respectively. Correspondingly, the number of items published in APSR precipitously declined. AJPS, which assumed its current name in 1973, was initially published as the Midwest Journal of Political Science in 1957. A decade later, two journals emerged that focused on comparative politics—CPS and CP, both in 1968; BJPS began publication in 1971. As figure 1 indicates, the publication of abstracts that summarized the content of articles occurred much later, accelerating in the 1970s and 1980s. Previously, “abstracts” provided by the publisher consisted of preliminary content.

Figure 1

Number of Citations by Journal

TRENDS IN POLITICAL SCIENCE RESEARCH

The turn of the twentieth century marked the birth of political science in the United States: the American Political Science Association was founded in 1903; its flagship publication, American Political Science Review, was first published in 1906; and the Social Science Research Council was formed in 1923. In addition to the many bulletins on political functions, much of the research published in APSR in the first few decades was concerned primarily with discussing American political institutions. During the first 20 years of research published in APSR, roughly 1 in 10 articles concerned Congress, whereas about 1 in 20 concerned the courts. Following the end of World War II, epistemological change occurred in the form of behavioralism, a philosophy influenced by sociology that called for broadening political science to include the study of informal procedures and behaviors (Dryzek Reference Dryzek2006). As Munck (Reference Munck, Munck and Snyder2007) noted, the development of a behavioral focus began in American politics as early as the 1920s, when scholars including Charles Merriam opposed historical and descriptive approaches to political science; however, comparative politics was not affected until after the war.

During the 1950s and 1960s, behavioralism was concerned with constructing an overarching theory of politics based on observations about the relationships among social institutions, embodied by a prominent theory in sociology called “structural functionalism” (Boix and Stokes Reference Boix, Stokes, Boix and Stokes2007). ^{Footnote 5} Behavioralism also encouraged a more scientific approach to theory and methods called “positivism” (Lane Reference Lane1996). The result was a greater emphasis on cross-national statistics and an expanded focus beyond big countries in Europe (Munck Reference Munck, Munck and Snyder2007). An example of the culmination of behavioralism is in Almond and Verba (Reference Almond and Verba1963), who analyzed mass attitudes and values across five countries to explain the role of culture in supporting democratic institutions (Sabetti Reference Sabetti, Boix and Stokes2007).

As criticism of the behavioralist approach to political science grew, a new generation began to reshape the field and introduce a variety of new ideas. Although the research that comprised the “post-behavioralist” period was not unified, Munck (Reference Munck, Munck and Snyder2007) argued that it was characterized by a focus on mid-range theory—that is, developing and testing hypotheses that did not aim to identify an overarching causal factor—and an emphasis on political determinants and instruments of politics. The methods associated with behavioralism continued, however; comparativists remained wedded to qualitative and small-N studies, due in part to issues of data availability. The result, therefore, was a division among comparativists over the relative merits of qualitative and quantitative research and the gradual distancing of comparative politics research from American politics (Munck Reference Munck, Munck and Snyder2007).

In the late 1980s and 1990s, comparative politics research took another turn that Munck (Reference Munck, Munck and Snyder2007) called the “second scientific revolution,” so named because behavioralism was also concerned with the scientific method. Like the behavioralists of the 1950s and 1960s, advocates of this new approach shared the goal of constructing unified theories, but they drew on concepts and tools from economics. ^{Footnote 6} Figure 2 compares references that might appear to characterize the “behavioral” and “second scientific” revolutions in comparative politics. Between 1950 and 1990, the predominant focus was on political behavior, which was surpassed after 1990 by a change in focus to political institutions. The rise of institutionalism in political science entailed the contemporaneous emergence of three different strands of “institutionalism”—historical, rational choice, and sociological—each with a slightly different take on the role of political institutions (Hall and Taylor Reference Hall and Taylor1996). The 2000s also witnessed the extension of the institutional “paradigm” to the study of comparative authoritarianism (Pepinsky Reference Pepinsky2014). A comparison of all journal citations to those in comparative politics–specific journals (i.e., CP, CPS, and WP) is quite similar (see the online appendix).

Figure 2

Article References to Institutions and Behavior

The research that comprised the behavioral revolution in comparative politics tended to treat the role of the state as a “black box,” which prompted a renewed effort to “bring the state back in” in the 1970s (Evans, Rueschemeyer, and Skocpol Reference Evans, Rueschemeyer and Skocpol1985; Munck Reference Munck, Munck and Snyder2007). As figure 3 shows, the second scientific revolution constituted a renewed interest in the state and a shift in focus to political institutions, as well as the adoption of frameworks common to economics. ^{Footnote 7} The trends are roughly the same among comparative politics journals, consistent with the timing of their publication (see the online appendix). Concurrent with increasing references to the state, scholars adopted game-theoretic and rational-choice models that were commonplace in economics—the growth of which was notable in the 1980s and 1990s. The late 1960s also witnessed an increasing interest in political methodology and the use of data to test theory (Boix and Stokes Reference Boix, Stokes, Boix and Stokes2007; Lichbach and Zuckerman Reference Lichbach, Zuckerman, Irving Lichbach and Zuckerman2009). ^{Footnote 8}

Figure 3

Article References to the State, Rational Choice, and Formal Models

According to Dalton (Reference Dalton2000), models of cleavage- and partisan-based voting began to be challenged in the 1980s. A wave of democratization in the developing world that began in the mid-1970s also spurred an interest in the linkages between citizens and elites (Kitschelt Reference Kitschelt2000). Questions stemming from the growth in electoral research included the effects of electoral rules on the size and number of parties, the motivations and behavior of candidates in elections, and the determinants of vote choice. Along with the increasing focus on elections and electoral choice, there was a shift in focus to individual voters and candidates, as evidenced in figure 4. Thus, as scholars moved away from thinking about group-based models of voting, they turned their attention to greater theorizing about the effect of elections on individual decision making.

Figure 4

Article References to Issues, Candidates, and Elections

In the comparative politics journals that Munck and Snyder (Reference Munck and Snyder2007) analyzed between 1989 and 2004, research on economic policy and reform constituted the largest share of articles. As a result, figure 4 also illustrates references to specific policies and issues, demonstrating a sharp increase that occurred in the 1970s. The uptick in research on policy and reform resulted from growing concern about making political science research more accessible to policy makers. The Policy Studies Organization emerged in the early 1970s, and the first issue of the Policy Studies Journal was published in 1972.

Given the increased attention given to individuals, the turn of the twenty-first century also was characterized by an increase in research on identity politics, as exemplified in figure 5. Although nowhere near the same levels as broader subjects—including behavior and institutions, the state, and policy and elections—aspects of identity (e.g., gender, ethnicity, and religion) were most prominent in the 2000s. Specifically, the roles of ethnicity—and identity, more broadly—became more salient as scholars began to apply constructivist arguments to explain outcomes related to conflict and political economy, which treats identity as “fluid and endogenous to a set of social, economic, and political processes” (Chandra Reference Chandra2001, 7). Ethnicity was a topic that concerned roughly 1 in 10 articles in 2006, which also was true of identity in 2013. As Munck and Snyder (Reference Munck and Snyder2007) noted, however, religious identity received far less attention, which also is true of research on gender. Research that included references to females or involved the topic of gender represented roughly 1 in 20 articles in 2006, whereas about 4% of references in 2013 concerned religion. Nevertheless, the previous 40 years have witnessed more research than ever before that addresses issues related to identity politics, which is concurrent with growth in the focus on individuals and policies.

Figure 5

Article References to Identity Politics

SUPPLEMENTARY MATERIAL

To view supplementary material for this article, please visit https://doi.org/10.1017/S104909651700110X.