British politics is a discipline with a lot to commend it. It is a mainstay of UK higher education via the many politics degrees offered at UK universities, and there are a plethora of academics doing interesting and impactful research on a wide variety of British politics-related topics. British politics scholars have also in recent years been at the forefront of important societal debates over such things as the rise of populism, the emergence of digital democracies and, especially, Brexit in all of its various dimensions (see, for example, Ashcroft and Bevir 2021; Gamble 2019; Goodwin and Milazzo 2017; Jennings and Stoker 2017; Jessop 2017; Sobolewska and Ford 2020; and, Thompson 2017). Forty percent of the impact case studies submitted to the 2021 Research Excellence Framework had a clear British politics bent, which suggests that British politics research continues to produce a broad range of tangible benefits for society.

However, the discipline has also faced a number of challenges in recent years. Both Kerr and Kettell (2006) and Beech (2012) have written about the tendency of British politics to be sidelined, first by the academic vogue in the 1990s and 2000s for studies of ‘globalisation’ and then, institutionally, by a Research Excellence Framework premised on an ‘arbitrary and hierarchical’ distinction between nationally and internationally renowned publications, which is bound to place disciplines such as British politics, with a clear national focus, at a disadvantage.

British politics as a discipline has also been criticised for having been too reliant on the theoretical lens provided by the Westminster model to structure its field of inquiry. One of the consequences of this has been to give the discipline an unduly narrow focus on the institutions of British central government (Gamble 1990; Kerr and Kettell 2006). This has traditionally gone hand-in-hand with a relative disinterest in theory and methods, with many authors instead opting for highly descriptive accounts of ‘core institutions, actors and processes’ (Kerr and Kettell 2006, p. 15). This over-reliance on the Westminster model has also tended to make British politics scholars less interested in international comparison and cross-disciplinary working (Kerr and Kettell 2006; McAnulla 2006; Marsh 2012).

Some of these problems were partly addressed by the Westminster model being supplanted as the dominant model of British politics by the governance or ‘differentiated polity’ model from the 2000s on (Marsh 2012). This approach to the study of British politics shifted attention away from the institutions of British central government—depicting them as ‘hollowed-out’ and ‘resource dependent’—and on to diffuse policy networks (see Bevir and Rhodes 2003, 2004, 2006). However, as noted by Marsh (2008; see also Marsh 2012 and Marsh et al 2003) in the development of his rival ‘asymmetric power’ model, the governance approach not only overstated the degree to which hierarchy had been replaced by network governance, but also crucially failed to incorporate into its analysis asymmetries of power in civil society, in particular relating to class, gender and ethnicity. In other words, where there was once a flourishing literature grappling with class conflict and its containment in Britain from the 1960s on (Gamble 1990, p. 410), such concerns started to become marginalised in the 1990s as attention shifted to the dynamics of multi-level governance. Few authors have taken the opportunity of Brexit to try to revive the discipline’s interest in Britain’s class politics, albeit with the notable exception of Evans and Tilley (2017).

Others such as Randall (2012, p. 17; see also BJPIR 2022, Akram 2023) have cited as ‘grounds for concern’ that British politics as a discipline has repeatedly shown blind spots for a range of issues of crucial importance in contemporary Britain, including the politics of age, ethnicity and the environment, that it has not fully capitalised on insights from feminist scholarship, and that when it has engaged with some of these topics it has often done so in a limited, electoral outcome-focused way (Randall 2012). Furthermore, these problems are said to be rooted in a lack of representativeness within the discipline and, in particular, ‘the virtual absence of Black and Minority Ethnic (BME) voices and perspectives in the contemporary study of British politics’ (Randall 2012, p. 24; see also Awesti et al. 2016).

All of this begs the question, what is British politics scholarship for? Certainly, as Beech (2012, p. 14) notes, British politics merits study on the ground that ‘Britain… is a highly significant nation-state’ because of the decisive role it played in world politics throughout the nineteenth and twentieth centuries, and because it remains an important international actor in the economic, security and development fields (even if this may have been diminished by Brexit). We might add to this that Britain is worth studying as a sui generis country case given the distinctive history of the Westminster model, the many quirks of which were on full display in the parliamentary battles over Brexit between 2017 and 2019. However, there is perhaps an additional justification. Implicit in much of the critical scholarship reflecting on the discipline is a view of British politics as having a normative or even transformative purpose. Despite the more intense focus on ‘employability’ in UK higher education in recent years, including in relation to the design of politics degrees, and the increased instrumentalization of knowledge production in social science research generally, it remains true that a fundamental purpose of the study of politics is the development of an understanding of the political world that can be used to find solutions to important social and political problems.

In other words, we know that many of the most urgent problems we face as a society are rooted in politics, or are at least politically mediated, and that while some of them—such as climate change, the security dilemmas created by global reordering, and the rise of populism—can only really be grasped in proper international context, many of them, including the iniquitous legacy of the pandemic, the potential break-up of the UK, and the dysfunction of the British economy, are perhaps best understood from a British politics perspective—albeit one that engages with a wider variety of social and political theories, and which seeks to learn from relevant international experience. In other words, if Flinders and Pal (2020, p. 273) are correct that the normative stance of the discipline of politics is ‘the restraint of political and social power by considerations of justice in the service of creating political orders that permit and encourage human flourishing’, then British politics has an important contribution to make to that.

Post-2008 financial crisis British politics, and especially the past decade, has been especially changeable, from the Scottish independence referendum, to the Brexit vote and the travails of the subsequent withdrawal process; to Covid lockdowns, and the more recent mounting sense of crisis driven by the post-Brexit economic malaise, the UK’s dysfunctional public services, industrial strife the likes of which has not been seen since the 1970s, and near unprecedented turmoil in the core executive, with four prime ministers in five years. Likewise, in the academy, the events of 2020 have led to a keener focus on ‘decolonising’ politics and other disciplines within UK universities (Begum and Saini 2019; Saini and Begum 2020; Chauhan et al 2020).

If British politics is to engage productively with the challenges of the moment, then we need a better understanding of the current state of the discipline. This article aims to provide such an understanding using network analysis to create a conceptual map of the discipline showing its key focuses as well as what it overlooks. It aims to shed light on the what of British politics scholarship by identifying the various research niches that make up the discipline. In doing this, it will hopefully allow for more meaningful self-reflection on the part of British politics scholars and help identify productive new avenues of research. In particular, it presents evidence that there is still work to do to achieve the laudable aims set out by Kerr and Kettell (2006) at the inception of this journal, especially in terms of fostering greater disciplinary cohesion in British politics research, and in meeting the political challenges of today by bringing together historical and contemporary analyses. Then, in terms of the who of British politics scholarship, we explore the role of gender in structuring patterns of research within the discipline.

In the next section we explore what network analysis, the method used here, can potentially tell us about the discipline of British politics. Then in subsequent sections we present the findings of our network analysis of British Politics, geared towards answering our main research question, What kind of discipline is British politics? The core focus in this respect is identifying and understanding the clusters of researchers forming distinct communities within the broad field of British politics scholarship, based on citation patterns. Then we reflect on the significance of these findings for the discipline going forward, arguing that the failure to properly politicise the Westminster model has left the discipline of British politics without some of the tools it needs to adequately address the big challenges we face as a society today.

What can network analysis tell us about the discipline of British politics?

As valuable as the recent literature on the discipline of British politics, discussed in the previous section, is, what is needed is an up-to-date, empirical accounting of the claims it makes. Network analysis has the potential to provide such an accounting. At base, network analysis is a research method used for ‘measuring and analyzing the structural properties of networks of interdependent dyadic relationships’ (Steketee et al 2015, p. 461). In other words, it uses a suite of graph theoretical mathematical techniques alongside detailed visual representation of networks to highlight the salient patterns they contain (Scott 2017).

Network analysis has a wide range of applications across both the natural and social sciences. It has been used to good effect in the fields of psychology (Griggs and Collisson 2013), economics (Bodenhorn 2003); higher education pedagogy (Calma and Davies 2017, 2015), and medicine (Christakis and Fowler 2007; Liu et al. 2010; Xian and Madhavan 2013; Sommer and Rappel-Kroyzer 2022). There have been a handful of noteworthy applications of the method in politics scholarship (Asher et al. 2019; van Esch et al. 2016; Wallaschek et al 2022) and, relatedly, international political economy (Seabrooke and Young 2017), but in general terms it remains underutilised, and there has to date been no application of network analysis specifically to study of British politics as a discipline—a gap this study aims to fill.

This disinterest in network analysis is unfortunate because the value of the method stems from the fact that networks are at the core of our social and political existence and, bearing in mind the inherent interdependency of observations in politics research, it stands to reason that they require a specialised research method of their own (Steketee et al 2015, p. 461). Furthermore, not only is it valid, but also it is to be encouraged that academic disciplines demonstrate a self-reflexive approach to their own enterprise. Every academic discipline needs the space to work out what its positive contribution is, and a means of determining whether it is actually making that contribution. Thanks to the dyadic nature of the citation-based relationships between research articles, network analysis provides scope for doing just that.

Network analysis’ quantitative tools can be categorised into node- and network-level measures. Node-level measures centre around a variety of ways of determining the importance or ‘centrality’ of particular nodes. They include: degree centrality, which refers to the number of connections a node has; betweenness centrality, which refers to the number of pairs of nodes a given node connects (and is useful in finding places where a network would break off if a node with high betweenness was removed); eigenvector centrality, which refers to how well-connected a node is to other well-connected nodes; and closeness centrality, which is the average distance from a given node to all other nodes in a network. Meanwhile, at the network level there are measures such as network density, which refers to the ratio of the actual number of connections with a network to the maximum possible number of connections in that same network, and network diameter, which refers to the largest distance between two nodes in a network. These measures can be useful for the purposes of comparing networks, or different states of the same network over time. Additionally, at the network level attention focuses on the number, size and nature of subgroups within a network.

These techniques allow us to visually map out the structure of this journal as a proxy for the broader discipline, and and to empirically verify claims about what it focuses on and its (lack of) representativeness. First and foremost, our aim is to use the techniques of network analysis to detect, and then rank in terms of size and centrality, distinct communities of researchers operating within the broader field of British politics research as a means of determining what its key focuses are. This speaks to our concern with the what of British politics research. Then, shifting our focus on to the who of British politics research, we explore which individual pieces of research are the most important or central in light of stark gender imbalances in British politics research.

In terms of the practicalities of the research, the analysis was carried out in the network analysis software, Gephi, which visualises networks as comprising a multitude of ‘nodes’ (small dots) representing individual network actors, and ‘edges’ (lines between the dots) representing connections between the nodes. Our visual map of the journal was developed by making the nodes in the network graph represent outputs published in British Politics, with the edges representing citations between articles. The key functionality of Gephi is the various algorithms it uses to give network graphs a meaningful structure, most of which aim to bring together highly connected nodes, and to space apart unconnected ones, in ways that make the network more easily readable.

Gephi can also be used to algorithmically detect ‘modularity classes’ within a given network. ‘Modularity’ refers to a network’s tendency to cluster into groups, which can be usefully visually represented within the network graph using colour coding. One of the key objectives of the study was to determine what the various sub-fields or niches are within British politics research, and to understand how they relate to one another. Most scholars of British politics would have an intuitive sense of what British politics as a discipline is like and what the key dividing lines are, but a major advantage of network analysis is that these groups are detected automatically by an algorithm and, therefore, without the imposition of a prior conceptual frame.

However, there are several methodological issues that need to be borne in mind when using network analysis in this way. Firstly, there is clearly an in-built bias against more recent research because we can expect such papers to have fewer average citations per article. This is simply because it typically takes a while for a paper to find its audience. However, we have tried to mitigate this problem by excluding from our analysis articles published in the last 12 months (bringing us up to the start of 2023, at the time of writing). We could have been even stricter with this cut-off, but for the need to balance it against the rival concern of making newer articles part of our picture depicting the current state of British politics. Had we gone back several years, we might have missed recent salient trends in British politics scholarship. Furthermore, there is a ‘life cycle’ to article citation counts in virtually every discipline, including politics. Annual citation counts tend to rise for several years (even for articles that eventually come to be very highly cited it can take a year or two for authors to find out about and incorporate them into their own published research), then peak (depending on the discipline, this usually takes somewhere between 3 and 5 years), and then tail off precipitously until there are almost no new citations (Galiani and Gálvez 2017). This means that the citation bias in favour of older articles is less pronounced than would be the case if there was no journal citation life cycle and citation counts tended to increase linearly year-on-year. Some articles—usually ones that come to be paradigm-defining in theoretical terms—do exhibit this quality, and this helps to explain why most of the top cited articles listed below (see Table 1) are older than ten years, but these are fairly exceptional. Another potential solution is to only use as part of the analysis ‘highly cited’ articles that have more than the mean number of citations per year, or some other such threshold. However, the downside to this is that it would effectively cut our dataset in half, which is a problem given that British Politics already has fewer than 500 published outputs to date.

Table 1 Top 10 British Politics (journal) articles by citations
Full size table

A second methodological issue to acknowledge, and a more fundamental one, is that although the statistical measures described above are useful, the analysis presented later on in this article does depend to a large extent on the authors’ interpretation of the structure of the network graphs. In particular, the software algorithmically detects and colour codes distinct communities based on the links between nodes (in our case, journal article citations), but is not capable of determining the meaning of each of the groups, instead merely giving them nondescript numeric labels. It was therefore left to the authors to explore the groups and to interpret the meaning of the data, and to give the communities meaningful names based on our analysis of the individual articles they contained.

Thirdly, this study is limited in that it uses the published outputs of a single academic journal as a proxy for an entire discipline. This is not necessarily a fatal limitation—indeed, there are many good bibliometric analyses that take this approach (see, for example, Calma and Davies 2015 and Calma and Davies 2017). It is also justified based on the inapplicability of traditional quantitative sampling methods in network analysis. Given the infeasibility of building a dataset that encompasses virtually the entire discipline, it is better to analyse the entire output of a single particularly representative journal than aim to sample a small proportion of outputs across a selection of journals (Scott 2017, p. 51). Furthermore, even though there are several other journals that frequently publish British politics-related articles—The Political Quarterly, The British Journal of Politics and International Relations, and Political Studies, to name a few—all of these are general politics and international relations journals, albeit from a distinctly British perspective. British Politics, on the other hand, is really the only journal in existence with an explicit substantive focus on ‘British politics’. That said, focusing on a single journal could well be a source of bias for other reasons. It is not possible to state with absolute certainty that all of the findings set out below truly speak to the nature of the discipline rather than the idiosyncrasies of the journal British Politics, and it clearly is the case that particular communities of scholars are selective in terms of where they choose to publish. For example, one of the aims of the research is to determine whether or not race and ethnicity as structured inequalities in British politics receive their due attention. Historically the most prominent voices on these topics have been a group of scholars associated first with the Centre for Contemporary Cultural Studies and then the journals New Left Review and Soundings (see, for example, Hall 1979, 1985, 2003, 2011). For this reason, confirming our findings will require a more expansive study in future.

Procedurally, how the analysis worked was to build a spreadsheet in MS Excel containing all of the nodes for the network graph and their corresponding edges. With the nodes being journal articles, these comprised every article published in British Politics to date (totalling 475 articles) and every article (some 17,500 of them altogether) that cited them. These were extracted using the online Scopus database. The nodes Excel file contained information on author, title, year of publication, issue number, source title, number of citations, and gender, as well as a unique ID for each article. The edges Excel file contained information on source, target and ‘Type’ (meaning, directed or undirected). Where any of this data had been omitted by Scopus (e.g., the gender of the lead author), it was manually added by the authors. These files were then imported into Gephi as the raw data for the network graph. The network graph was given structure using the layout algorithm ‘ForceAtlas 2’, designed to bring together highly connected nodes and to space apart unconnected ones while maintaining visual readability. This layout algorithm is also appropriate for large networks like ours. The network graph was then further modified for readability, including indicating citation counts with node sizing, and modularity classes and gender with colour coding. These modifications formed the basis of the different versions of the graphs shown in the figures below. A statistical analysis looking at modularity classes, betweenness centrality and average degree was performed in Gephi and, where necessary (in the case of, for example, average degree by modularity class), in Excel. After the modularity classes were produced in Gephi they were subjected to an interpretive analysis by the authors to attach appropriately representative labels.

What kind of discipline is British politics?

Figures 1 and 2 visually map out the structure of British politics research, as derived from our network analysis. They illustrate the continued influence of the Westminster model, as well as several other salient features of the discipline. Both figures centre on the published output of British Politics since 2006—with the nodes representing articles published within the journal and edges representing citations, and with node size varying according to degree centrality and colour coding representing different modularity classes—but Fig. 2 is a larger network that also encompasses citing texts from outside the journal. It is therefore a much larger network. As we will see, it has some utility in terms of understanding links between British politics and other parts of the discipline of politics, as well as other disciplines. However, Fig. 1 is more intelligible. What it most starkly illustrates via the colour coded modularity groups is that the discipline is fragmented into a large number of distinct research niches, and that these niches speak to core concerns of the Westminster model.

Fig. 1
figure 1

British Politics (journal) modularity classes

Full size image
Fig. 2
figure 2

British Politics (journal) and citing articles modularity classes

Full size image

As noted above, these groups were automatically detected by an algorithm within Gephi based on the number of close connections between articles. Labels were then applied by the authors after an interpretive analysis of the algorithmically detected groups. The figures also include statistics on the size of the various groups, both in terms of the overall number of articles they each comprise and average degree centrality. Figure 1 shows that the largest research niches that emerged from this analysis focus on: conservatism and Conservative Party politics; the British political tradition; party leadership elections; the policy process (including literature on depoliticisation); British politics as a discipline; and, the fall of New Labour (especially its handling of the 2008 financial crisis). Each of these accounted for at least 5% of the entire corpus.

However, it is interesting to note that the largest groups measured in terms of article counts were not necessarily the ones with the highest average degree centrality—meaning that there was a disjuncture between the most common kinds of articles submitted to and published by the journal, and the kinds of articles other authors engaged with the most. Using this metric and looking at the most ‘central’ groups measured in terms of average degree centrality, we find that although the same groups make it into the top ten in terms of both size and centrality, their rank order differs markedly. Although studies of conservatism and the Conservative Party continue to occupy the top spot, the pecking order below is different, with leadership elections, British political rhetoric, the British political tradition, and British politics as a discipline occupying the top five slots, and with Thatcherism and the study of political campaigning and communications following closely behind.

Nevertheless, in terms of the geography of the graph, and bearing in mind that the algorithm used to construct the network graph is designed to bring together highly connected nodes and to space apart unconnected ones (which implies that cognate sub-disciplines tend to be located close to one another), what is most clearly evident is that the overall structure of British politics as a discipline reflects the continued outsized influence of the Westminster model. This overall structure of the network can be summarised thusly: firstly, we have one large cluster of party politics-focused niches (primarily the modularity groups for conservatism and leadership elections) in the upper left quadrant of the graph; another large cluster of political economy-focused niches (New Labour, Thatcherism, rhetoric) in the upper right quadrant; and, a third large cluster of theory-focused niches (British politics as discipline, the British political tradition, depoliticisation and the policy process) in the lower half of the graph.

To a limited extent, this network graph contradicts some of the claims made in the literature on the discipline of British politics discussed above. Much of the political economy-focused cluster takes a broader view of British politics, drawing on a variety of social and political theories to try to make sense of ideological trends in British politics. Authors inspired by the ‘statecraft’ approach first popularised by Jim Bulpitt, and Stuart Hall’s cultural analyses of Thatcherism, play an important role here. However, much of this cluster also remains preoccupied with the institutions making up the Westminster model, and in particular the prime minister and chancellor. Similarly, although much of the literature making up the rhetoric-focused cluster draws on diverse theoretical influences ranging from classical political theory to discourse theory, in terms of its substantive empirical focus, it remains almost without exception the speeches of the prime minister.

Meanwhile, although it is to be welcomed that there is a clearly discernible theory-focused set of research niches, this engagement with theory operates within narrow bounds, and in some cases is outright misleading. For example, there has in recent years been a growing literature on British politics as a discipline, but a large proportion of this has focused straightforwardly on the impact of UK government higher education policy (for example, the Research Excellence Framework) rather than critically reflecting on the nature and purposes of the discipline. Similarly, a large chunk of this sector is made up of ‘middle range’ theories focused exclusively on the policy process—which again reflects a preoccupation with Westminster.

Figure 2 serves to triangulate the findings depicted in Fig. 1 because even with the expanded dataset most of the largest groups were the same or very similar. These were: governance (including many interpretivist analyses of British politics and a handful of accounts of the British political tradition); conservatism and Conservative Party politics (with the bulk of the literature referring to the coalition government and austerity); depoliticisation; political leadership (including leadership elections); and opposition leaders (with many of these articles focusing on Corbynism).

Overall, this analysis of the largest research niches in the discipline of British politics speaks to a discipline that is preoccupied with two main things: the institutions of British central government and party politics. Meanwhile, although the two are broadly very similar, considering the differences between the British Politics articles-only graph (Fig. 1) and the larger graph that also includes citing articles from outside the journal (Fig. 2), the main thing that stands out is the cluster of research around climate change and environmental politics in the latter, and the greater focus on British politics as a discipline in the former. This disjuncture implies that British politics scholars are relatively less interested in the politics of climate change than other parts of the discipline of politics.

Turning our attention to node-level measures, Tables 1, 2, and 3 each present a distinct method for ranking the importance of the articles in the dataset. These are based on straightforward citation counts, degree centrality and betweenness centrality. Although, as expected, many of the same articles are present in each of the lists, there were some notable differences. As noted above, betweenness centrality is a way of measuring centrality premised on counting the number of shortest paths between pairs of nodes in a network that a particular node lies on. In other words, the node that lies on the most shortest paths between pairs of nodes within a network will have the highest betweenness centrality. However, this measure can be problematic for a variety of reasons. For example, articles that have either no in-degrees (in our case, citations) or no out-degrees (references) by definition have zero betweenness centrality—that is, they cannot perform a bridging function between two nodes because they are connected, at most, in a series of two-way relationships. This can make important articles within the network seem unimportant. Nevertheless, this measure can be used to add more nuance to the blunt approaches of citation and degree counts. The top ranked articles by this measure are presented in Table 3. There are clearly some commonalities among these articles, which help to explain their success in bridging different parts of the network. The vast majority of the articles focused on Conservative Party politics or—failing that—New Labour. This speaks to British politics’ preoccupation with the philosophy and political economy of the governing party. Other topics that feature prominently were leader selection and Europe broadly (including Brexit). Meanwhile, in terms of theory and methods, it is noteworthy that virtually all of the articles in this list had a more general, cross-cutting, rather than policy area specific, focus, and that they served a useful purpose in terms of providing other authors with theoretical and conceptual tools that could be put to good use across a range of research niches. This is further illustrated in Fig. 3, which represents the betweenness centrality of nodes using a colour gradient, with darker nodes being more central to the network.

Table 2 Top 10 British Politics (journal) articles by degree centrality
Full size table
Table 3 Top 10 British Politics (journal) articles by betweenness centrality
Full size table
Fig. 3
figure 3

British Politics (journal) betweenness centrality

Full size image

British politics research and gender

In terms of who publishes in the journal British Politics, and what that says about the discipline as a whole, there is a notable gender imbalance: just 20% of articles had a woman as the first author (strangely, although there were quite a few mixed gender co-authored articles with a woman as the lead author, there were none with a man as lead author). Until the Covid-19 pandemic the historical trend was for the proportion of women first authors to increase, such that in the last year before the pandemic 38 percent of first authors were women. However, that cratered to 25% in 2020 and 15 percent in 2021—which is barely any better than the 14 percent at the journal’s inception in 2006. Covid-19 is undoubtedly partly responsible here, with a much higher percentage of women academics reporting in surveys that they took on unanticipated childcare and elder care because of the pandemic, greater anxiety about a family member contracting Covid, and inability to concentrate on research for reasons related to the pandemic (Johnson et al 2021).

The authors with the most articles in the journal, excluding editorials and other articles by editors, were Peter Dorey, Timothy Heppell and Paul Cairney (with 11, 10 and 9 solely or co-authored articles, respectively). All but one of the ten authors with the most publications were men. There was also a disproportionate number of articles from Russell Group universities, with all but one of the top 10 best represented institutions coming from among this grouping. In terms of nationality, perhaps inevitably, the vast majority (85%) of articles were from UK-based authors, with the US (6%) and Australia (5%) coming second and third. In terms of citations, the top cited article overall was Flinders and Buller (2006). None of the ten most cited articles were written, or even co-authored, by a woman. Likewise, only one of the top ten articles by degree centrality was written by a woman.

A closer consideration of the gendered nature of British politics research networks is also revealing. Figure 4 is essentially the same network graph as Fig. 1, but colour-coded by gender (with men coded red and women coded green). It helps to explain the gender imbalance discussed above in terms of lead authors of journal articles. Clearly there is a large gender gap in British politics research. There are far fewer women than men among authors featured in the journal, affirming Williams et al’s (2015) finding that women are less likely to be published as the sole or lead author in top British politics and international relations journals (see also Bates et al. 2012).

Fig. 4
figure 4

British Politics (journal) gender

Full size image
Fig. 5
figure 5

Percentage of British Politics articles with woman as first author by year. The percentage figures are based on the gender of an article’s lead author. The lead author’s gender was coded based on our list of article authors’ names

Full size image
Fig. 6
figure 6

Women as a percentage of first authors within modularity class

Full size image

The network analysis presents some additional interesting findings though. Firstly, almost all of the largest nodes on the network graph—meaning the ones with the most degrees or citations—are written by men, reflecting their greater centrality within the discipline. Secondly, it is particularly noteworthy that although women are represented in all of the major sub-disciplines of British politics research, there are some vast disproportionalities, with women scholars being best represented in research niches focusing on political campaigning and communication, and political rhetoric (although even in these groups women do not constitute a majority of first named authors). Conversely, women were exceptionally poorly represented among studies of Thatcherism, party leader selection and depoliticisation, as outlined in Fig. 4. This may reflect the fact that the former are where much of the available research funding is concentrated, the targeting of which is one of the few ways in which female scholars can counteract structural inequalities in academia. A more wide-ranging analysis incorporating a broader range of academic journals would tell us if women British politics scholars also tend to publish in journals focused more specifically on gender and politics (Figs. 5, 6).

Conclusion

What have we learned about British politics research that may help us answer our initial research questions? Those questions centred around the what and the who of British politics research. In terms of the what, we were interested in the overall structure of the discipline and, in particular, whether the various communities of researchers making it up focused on a sufficiently broad range of topics, encompassing the gaps previously identified by Randall (2012) and others. Furthermore, what can we say about the continued influence of the Westminster model? In that respect, we have learned that although certain parts of our analysis—the growing importance of research on climate and the UK’s territorial politics in particular—contradict the narrative of Westminster model dominance, for the most part the discipline continues to be organised around this perspective, with the latter’s key actors, institutions and processes remaining a focal point. This is not to dismiss the usefulness of the concept or ‘inherited tradition’ of the Westminster model within interpretivist scholarship, bearing in mind the purchase it has on actual practitioners of British government (Flinders et al 2022; see Russell and Serban 2021 for a more critical assessment of the usefulness of the model in comparative political science). However, what it does call for is more careful consideration on the part of British politics scholars of how their research might be influenced by the Westminster model as something that directs their attention towards certain topics and away from others, and for more attention to be paid to issues of race, ethnicity, gender, class and the politics of age—or, in other words, for the ‘politicisation’ of the Westminster model.

In terms of the who of British politics research, our aims were more modest, given that we only explored a single variable—albeit the crucially important variable of gender—in any real depth. That said, we learned some important things: most notably, that there is a significant gender imbalance in British politics research, and that although the discipline is changing, both in terms of women’s representation and, analytically, with a greater focus on gender politics, the direction of travel is not straightforward. In particular, there is stark evidence of these trends going into reverse with the onset of Covid-19.

Clearly, this study also leaves many questions unanswered. What we hope to achieve with it, though, more than anything else is to bring more British politics scholars into debates around the nature and purposes of our common endeavour, and around how the discipline should develop over the coming years. For our part, our key goals going forward include: expanding the scope of this research to encompass a larger sample of British politics and politics research—are these lacunae being addressed in allied disciplines, or by politics scholars peripatetically interested in British politics?—to make us more certain of our findings and to give us a clearer picture of the nature of our discipline; exploring the discipline of British politics in all of its dimensions, going beyond looking at research on British politics to also encompass the way it is taught; and, lastly, expanding the scope of our analysis of the who of British politics research by bringing into the analysis the variables of age, class and ethnicity. Hopefully by doing that we will move closer to embodying the ideal of a discipline with a transformative purpose in relation to the wider society. We have in this paper tried to shed some light on what British politics is. What it should be is up to British politics scholars to decide.