Introduction
Western Europe entered the nineteenth century as a collection of nightwatchman states, more concerned with their imperial ambitions abroad than with steering, regulating, and disciplining their domestic population (Mann, Reference Mann2012, 408–411). Those imperial states had largely disappeared by the end of World War II, replaced by social states that were in the process of losing their colonial territories as they shifted their gaze inwards. At the core of this transformation was a renegotiation of the relationship between states and their subjects, especially regarding the purposes, instruments, and conditions for the deployment of state coercion toward its own people. Indeed, the public, specialized, and professionalized police forces that resulted from this renegotiation of state-society relations would ultimately become “the most immediately identifiable agents of the modern state” (Ansell and Lindvall, Reference Ansell and Lindvall2020, 63).
In this article, we apply computational text analysis to analyze how legislative debates about policing in the United Kingdom reflected broader political conflicts about centralization, routinization, and standardization in the exercise of state coercion. These processes likely motivated debates about policing at different points in time, driven by different sets of actors and concerns, and triggering political conflicts with different levels of intensity. To explore these issues, we need to go beyond a teleological reading of police reform – one solely based on the enaction of new legislation or commissions of inquiryFootnote 1 – so that we can capture the various moments, actors, concerns, and conflicts that shaped the legislative politics of policing even when such debates did not lead to concrete legislative outcomes.
To this end, we draw on a new corpus with the transcripts of all debates in both chambers of the British Parliament from 1803 to 1945. Those years witnessed the introduction, standardization, and centralization of a new model of policing that would ultimately be exported worldwide (Wall, Reference Wall1998). By applying computational text analysis techniques to almost 150 years of political debates, we are able to provide a comprehensive picture of when policing was actively discussed, what was the substantive content that shaped the conversation, who were the actors that dominated those debates, and how intense discursive conflict was in each of those instances. This allows us to both confirm established knowledge about the legislative history of policing in the United Kingdom, as well as unearth overlooked episodes, actors, and concerns that contributed to the politics of policing during its formative years.
Our analyses show that there were three distinct periods of extensive legislative discussion on the police. The first period (1830s to 1850s) was marked by the well-known institutional reforms that gave birth to modern police forces and later spread them across the country. These debates were dominated by Liberal MPs and versed around the relationship between the central government and the autonomy of local communities.
The second period (1880s–1900) has not received as much attention in the past. However, our analyses show that legislative debates on policing were more salient during the 1880s and 1890s than at any other time during the long nineteenth century. During this period, it was mainly Irish MPs who dominated parliamentary debates on policing. In the absence of local oversight bodies to hold the colonial police forces accountable,Footnote 2 Irish MPs used their office to raise complaints against police abuse, especially during the repression that followed the Land War of 1879–1882 and in the context of the Plan of Campaign in 1886–1891. Those debates point to the tensions generated by a modernizing state that sought to preserve imperial institutions abroad.
Finally, a third period (1900s–1940s) was marked by discussions around working conditions for police officers, wartime duties, and new societal challenges for the police. It was dominated by Labour and Conservative MPs and very much reflected the consolidation of a new kind of democratic state governed by bureaucratic routines.
These results indicate that the centralization and standardization of state coercion were not only at the center of legislative discussions about policing, but that they shaped those debates in different ways over time. Conflicts over the centralization and standardization of police forces were clearly the central concern of parliamentary debates during the first period of legislative activity that spread new models of policing across the provinces. They also emerged – albeit in a different guise – in the second period of discussions centered on the Irish tensions. Indeed, the parliamentary debates of the 1880s and 1890s highlighted the difficulties of reconciling the pressures towards the standardization of state coercion and the principles of political representation and the rule of law with imperial practices that subjected local communities to different coercive institutions. Whereas other communities in the United Kingdom had local control over police forces and local bodies to oversee abuses, Irish counties were policed by a militarized force that was ultimately accountable to London. However, contrary to other colonial territories, Ireland was at the time part of the Union and thus returned representatives to Parliament. This put Irish MPs in a unique position to use their offices in Parliament to demand accountability from what was seen as an occupying force by some of their constituents.
In what follows, we summarize the historical literature on the British police. We then describe the data and the methods. We structure our analyses in four sub-sections that focus on the timing of debates around policing (when), the concerns voiced in those debates (what), the actors that were involved (who), and the intensity of discursive conflict (how). We then zoom into the second period (1880s–1900s), which has remained at the margins of the historical literature on British policing. We conclude with some reflections on the relevance of our findings to the historiography of modern police forces and for social scientific research on the rise of the modern state.
The emergence and growth of modern police forces in the United Kingdom
The deployment of modern police forces in Europe during the late eighteenth and early nineteenth centuries was a major institutional innovation. The United Kingdom, together with France, was a pioneer in the establishment of “public, specialized, and professional” police forces aimed at regulating interpersonal relations and controlling public order through the (potential) use of force (Reference Bayley and Tilly1975, 23). They were meant to complement (and ultimately replace) the private and communal institutions that had been in place, and to repress social upheavals without the level of violence that characterized military forces (Bayley, Reference Bayley1990; Emsley, Reference Emsley2008; Mann, Reference Mann2012; Rawlings, Reference Rawlings2008; Reiner, Reference Reiner2000).
In the English case, the new model came to fruition with the establishment of the London Metropolitan Police (commonly referred to as “the Met”) in 1829 by Home Secretary Robert Peel. Although attempts to create modern police forces dated back at least to the unsuccessful Police Bill promoted by William Pitt in 1785 (and even to prior attempts to create a unified watch scheme in London), the 1829 Act was the first to implement a permanent and numerous civil force in the Metropolitan area, unrelated both organically and functionally to the army and directly accountable to the Home Secretary. It came amid fears of rising crime and disorder in a rapidly industrializing environment, which the long-standing watchmen and parish constables were deemed unable to tackle (Emsley, Reference Emsley1996; Finnane, Reference Finnane, Knepper and Johansen2016). The new police reflected the aspirations of the Liberal project, with a professionalized bureaucracy (regularly paid, uniformed, hierarchical, permanent, based on meritocratic rules of appointment and promotion) that was designed to guarantee the self-acclaimed English liberties. Policemen were armed only with a wooden baton and expected to act impartially and on principle, purportedly immune to the political ambitions of the dominant parties (Emsley, Reference Emsley1996). Their main responsibility was, allegedly, the prevention of crime, although their tasks ranged from apprehending individuals considered “idle and disorderly persons” to controlling undesirable activities such as careless driving or gaming (Emsley, Reference Emsley2008, 74).
This Metropolitan model became the blueprint for the standardization and bureaucratization of full-time and paid police forces in English and Welsh counties and boroughs during the next three decades, albeit in a much more decentralized fashion (Churchill, Reference Churchill2018; Emsley, Reference Emsley1996; Philips and Storch, Reference Philips and Storch1999; Wall, Reference Wall1998). Legislation was enacted to expand the full-time and professionalized police to the provinces, first on a voluntary basis (most notably with the 1835 Municipal Corporations Act and the 1839 County Police Act) and later based on mandatory clauses (culminating in the 1856 County and Borough Police Act, which required the establishment of professional police forces in all boroughs and counties). Metropolitan and provincial forces differed, however, in a crucial aspect, as the latter were accountable to the local authorities (Watch Committees in the case of boroughs, and magistrates – and later Standing Joint Committees – in the case of counties), while the former were responsible to the central government. This was the solution that legislators found to make the reforms compatible with a strong tradition of local autonomy and aversion to a strong centralized state. Even in the few cases where the central government had taken the initiative and appointed police forces under the authority of the Home Office (namely in Birmingham, Bolton, and Manchester, as a reaction to the threat posed by Chartism), control was handed over to local authorities after only three years of their implementation (Emsley, Reference Emsley2008).
These institutional innovations in the exercise of state coercion were part of larger processes of social and political transformation, as the expansion of the state into new policy areas also came with the growing bureaucratization of its offices. Administrative reforms established meritocratic guidelines for recruitment and delineating working conditions for public servants professionalized their roles and functions. Most importantly, interactions between the state and its subjects became increasingly routinized and standardized, governed by rules and guidelines, and abstracted from particularistic contexts (Cornell and Svensson, Reference Cornell and Svensson2023; Osborne, Reference Osborne1994; Thornhill, Reference Thornhill2008).
Despite the importance of these reforms, police historians are wary not to overemphasize the changes that took place at that time. Unlike some of the first historiographical accounts, which stressed the transition from a private and communal mode of policing to a public and efficient law-enforcement institution (e.g., Critchley, Reference Critchley1972; Reith, Reference Reith1943), most recent literature also highlights the continuities of new and old police forces. The degree of professionalization and the involvement of the state in the management of police forces were certainly relevant developments in the bureaucratization of policing, but this is better understood as a process of transformation and improvement upon previous police systems rather than a radical break with the past (Emsley, Reference Emsley1996; Finnane, Reference Finnane, Knepper and Johansen2016; Vogler, Reference Vogler, Jenkins and Rubin2022; Wall, Reference Wall1998). The new police corps were often filled with the very same constables and watchmen that guarded the streets prior to the reforms (Emsley, Reference Emsley1996; Wall, Reference Wall1998), and law enforcement responsibilities remained remarkably plural even after the new police forces were in place (Churchill, Reference Churchill2014, Reference Churchill2018).
In fact, other permanent and professionalized police forces had been established before the nineteenth-century wave of English police reforms. First, in London, the Bow Street Court established in 1749 a group of assistants aimed at aiding the magistrates in the detection and apprehension of suspected criminals (popularly known as Bow Street Runners), although it had a limited reach (at most with twelve men) and was a private institution (complemented with public subsidies). Other similar efforts, such as the private Thames River Police established in 1797 by dock owners to enforce new working practices, or even the publicly funded Horse Patrol, Foot Patrol, and Dismounted Horse Patrol, also predated the Met, even though they were limited in their scope (Emsley, Reference Emsley1996; Finnane, Reference Finnane, Knepper and Johansen2016; Rawlings, Reference Rawlings2008; Wall, Reference Wall1998). More relevant were the new police forces established first in Dublin in 1786 and later in several Scottish towns and cities, most notably in Glasgow in 1800 (Barrie, Reference Barrie2010).
The Irish police were particularly important to the history of British policing. Robert Peel, in his capacity as Chief Secretary for Ireland, created the Peace Preservation Force in 1814. This police force was merged with the County Constabulary in 1836 to form the Irish Constabulary (renamed as Royal Irish Constabulary in 1867 after the suppression of the Fenian rebellion). Based on a colonial and militarized pattern of police organization, the constabulary was spread through the territory in barracks embedded in local communities but apart from the civilian population. Unlike English provincial forces, it was firmly centralized under the control of Dublin Castle (Brogden, Reference Brogden1987). While the Irish Constabulary initially focused on political crime and agitation (Ellison and Smyth, Reference Ellison and Smyth2000), most of its interventions concerned minor infringements of regulations (Emsley, Reference Emsley1999b). It became a source of recruitment and training for colonial police forces across the British Empire, to the extent that by 1907 all colonial police officers of commissioned rank were required to receive training at the Royal Irish Constabulary station in Phoenix Park (Herlihy, Reference Herlihy1997). The creation of a gendarmery-style police force in Palestine in 1922 filled with former Royal Irish Constabulary officers (after it had been disbanded following the independence of Ireland) is also a clear illustration of the importance of colonial recruitment practices (Sinclair, Reference Sinclair2008).
The experiments with policing in Ireland were also a source of inspiration for modern police forces in other counties (Emsley, Reference Emsley2008), such as Gloucestershire and even for the Met itself. Some of the most important political actors that intervened in the design and implementation of the Met (e.g., Robert Peel, Arthur Wellesley, Charles Rowan, and Richard Mayne) drew from their experience with the Irish criminal justice system (Go Reference Go2023). According to Sinclair and Williams (Reference Sinclair and Williams2007), most police chiefs up to 1920 had had some experience in the colonial police forces, as part of a “cross-fertilization” between colonial and British policing.
Nevertheless, it is widely accepted that the establishment of the Met was a watershed event in the history of the British criminal justice system. Even if the 1829 Act did not establish an entirely new institution, none of the previous forces encapsulated the new trends in modern policing or had the same degree of projection (Finnane, Reference Finnane, Knepper and Johansen2016; Wall, Reference Wall1998). The Met did not only guide the design of provincial police forces in England and Wales, but it became an international reference for the so-called “state civilian” police, characterized as a civil public institution accountable to the central government (Emsley, Reference Emsley1999a). The adoption of modern police forces in English-speaking colonies and former colonies, such as the York (now Toronto) Police Force, the South Australia Police, and the New York Police was largely based on the English case (Ansell and Lindvall, Reference Ansell and Lindvall2020). The English model also influenced other countries. For instance, the establishment of the Guardia di Pubblica Sicurezza in Piedmont in 1852 was promoted by the Liberals along the lines of the British “bobby” as an alternative to the paramilitary Carabinieri, which they perceived as being controlled by aristocratic conservatives. Discussions in the Italian parliament during the 1860s even considered the deployment of local police forces modeled on the English provincial scheme (Emsley, Reference Emsley1999b).
The previous paragraphs show that the history of the British police has been studied extensively. However, research on the legislative politics of policing has largely focused on major episodes of reform, such as the 1829 creation of the Met, the 1839 County Police Act, the 1856 County and Borough Police Act, or the 1888 Local Government Act. These reforms standardized and routinized the exercise of coercion through increasingly professionalized police forces. We lack, however, a long-term analysis of how legislative attention to the police developed during the long nineteenth and early twentieth centuries, and how it related to the broader reconfiguration of the British state. A focus on key reforms inevitably overlooks periods of intense political activity around policing that did not culminate in concrete legislative outcomes. Such moments can nonetheless offer important insights into the political conflicts that surrounded the reconfiguration of the The British state and the transformations in the exercise of its coercive power. In the next section, we describe how computational text analysis can help us address these issues.
Methods and data
To evaluate how the politics of policing changed over the long run, we rely on a new dataset of all debates in both chambers of Parliament from 1803 to 1945 (Goenaga et al., Reference Goenaga, Pashley and Sabaté2024). The corpus expands on the work by Eggers and Spirling (Reference Eggers and Spirling2014), Rheault et al. (Reference Rheault, Beelen, Cochrane and Hirst2016), and the They Work for You project 2023, which collects parliamentary records for the more recent period, to create a single corpus, segmented at the speech level, for the House of Commons and the House of Lords. The dataset offers unique ID codes for each speech, debate (bounded discussions on the same topic with a title assigned by Hansard), parliamentary sitting, chamber, and name of the MP. Data on the MPs’ political parties were collected by Eggers and Spirling for the period 1830–1918. We have completed some missing observations and extended the party variable for the rest of the period with data from the EveryPolitician Project (N.d.). In all our analyses, we collapse the party variable into seven party families: Conservative, Liberal, Labour (including Lib-Lab MPs), Irish nationalists, Irish unionists, Liberal unionists, and others (which includes small parties such as the Scottish Prohibition Party, non-affiliated independent MPs, and major parties that were not very active during our period and topic of study, such as the Scottish National Party).
Our approach is primarily inductive and aimed toward discovery rather than testing (Grimmer et al. Reference Grimmer, Roberts and Stewart2022). If macro-historical processes of state centralization, routinization, and standardization motivated debates about policing, we expect this to show in the timing of those debates, the actors that were involved, the concerns that were voiced at the time, and the intensity of political conflicts around this issue. Hence, we structure our analyses around four main questions: (1) when was policing discussed; (2) what was the content of debates on policing; (2) who were the actors that participated in those debates; and (4) how was the tone of those discussions. These questions serve as heuristic tools to structure our analyses, as they refer to central questions in the historiography of policing: the periodization of reforms, the main issues driving those reforms, the actors involved, and the political conflicts that surrounded them. However, our approach enables us to explore these questions beyond specific instances of reform, even when political debates did not culminate in concrete legislative outputs.
First, to examine the timing of debates on policing, we combine supervised and semi-supervised approaches to assess the salience of this issue over time. We draw on the historiography of the British police to generate a dictionary of keywords related to policing (see Appendix A, Supplementary material).Footnote 3 We then use this dictionary in two different ways. For the first set of analyses, we generate a measure of “police salience” that divides the number of times any of our keywords was mentioned in a debate over the total number of words in that discussion. Debates that focused more on policing should thus display higher values. In addition, we count the number of debates per year that mentioned at least one of our police keywords in their title as a share of the total number of debates in that year. These two approaches present similar pictures of the timing of debates on the police in our corpus.
Second, to examine the content of debates on policing, we delimit our corpus to focus only on debates that included one of our police keywords in their titles. We then run Keyword Assisted Topic Models (KeyATMs) with decades as a covariate (Eshima et al., Reference Eshima, Imai and Sasaki2024). Compared to other topic-modeling approaches, KeyATMs allow the researcher to define ex-ante a topic by providing keywords that are highly discriminant of the topic of interest. The model then estimates the proportion of that topic in each document based on the presence of those words, as well as other words that tend to co-occur with them. As a result, these models allow us to identify debates on policing even if they do not repeatedly mention the keywords included in our dictionary. We specify the topics of substantive interest following a two-step approach. First, we run a frequency analysis to identify the most common words in the titles of the police debates (excluding our police keywords and other administrative and insubstantial words). We then complement that list with words referring to the most important issues discussed in the historical literature. Based on that list, we identify topics (and their related keywords) that we expect to find in the corpus. We started this analysis in 1830 because of the low police salience that we found prior to that date. In Appendix B (Supplementary material), we describe the process of topic selection in more detail.
Third, to capture which actors participated in debates on policing, we calculate the number of individual speeches by MPs for each party as a share of total speeches in debates on the police in each year. This gives us a sense of which political parties were more likely to dominate discussions at different points in time. Additionally, we run KeyATMs in which we include party as a covariate, in order to ascertain whether different political parties were more likely to evoke certain topics when discussing policing.
Fourth, we employ divisions and sentiment analysis to measure rhetorical conflict. For the former, we look at the number of divisions between 1835 and 1918, as well as the distribution of roll-call votes during this period, using data from Eggers and Spirling (Reference Eggers and Spirling2014). For the latter, we measure the intensity of contestation in debates on policing by calculating the sentiment score for each speech (Rinker, Reference Rinker2019) and then calculating the standard deviation across all speeches in a debate. The underlying intuition is that debates with greater variation in the sentiment score should be associated with higher levels of discursive contestation. We restrict the sample to debates with at least 5 speeches with more than 20 words each to avoid standard deviations based on very few or very short speeches.
When? Police salience in parliament
How did legislative attention to the police develop during the long nineteenth century? The top panel in Figure 1 explores this question by measuring police salience as the number of keywords related to the police over the total number of words in a debate. Each dot in the figure represents a debate (left axis) whereas the line shows the average level of police salience by year (right axis). The bottom panel in Figure 1 compares the average police salience with the average salience of the two policy realms that dominated legislative activity during the nineteenth century: military affairs and taxation (Goenaga and Sabaté, Reference Goenaga and Sabaté2020).
Figure 1. Police salience in parliamentary debates (1803–1945).
Notes: In the top panel, the dots represent police salience in each debate (measured as the number of keywords related to the police over the total number of keywords in a debate, left axis). The blue line represents the smoothed yearly average of police salience (right axis). The lower panel compares smoothed yearly averages of police, fiscal, and military salience. Fiscal and military salience from Goenaga and Sabaté (Reference Goenaga and Sabaté2020).
By focusing on the proportion of police-related keywords as a share of total words in a debate, Figure 1 captures mentions of the police but not necessarily debates on police reform per se. As David Wall (Reference Wall1998, 51) argues, it could be the case that policing was salient in parliament (for example, in discussions related to public order) but the police (and hence, police reform) were not. In Figure 2, we identify the number of debates that included our police keywords in their titles, and then we divide that number by the total number of debates in a year. The titles were assigned by Hansard and are not always very informative, so this measure is likely to underestimate the number of police debates. In other words, we can be confident that debates with these keywords were devoted to the police, but we might miss debates on police-related matters in which our keywords were not part of the titles.
Figure 2. Police salience by debate titles (1803–1945).
Note: Percentage of debates mentioning police keywords in their titles as a share of total number of debates in a year. The keywords used to identify police debates are: “police,” “policemen,” “policeman,” “constabulary,” “constable,” and “constables.”
Three interesting patterns emerge from these figures. First, our measures correctly capture well-known periods of police reform in the late 1830s, the 1840s, and the 1850s, when modern police forces were extended to the provinces. This is especially clear in Figure 2, where we observe two noticeable peaks that coincide with major police reforms – namely, the 1839 County Police Act and the 1856 County and Borough Police Act.Footnote 4 The former recommended the creation of full-time and paid police forces in counties, placing their control in the hands of the county magistrates, and it was preceded by commissions that recommended the improvement of the county forces, such as the 1839 Royal Commission on Establishing an Efficient Constabulary Force in the Counties of England and Wales.Footnote 5 However, the adoption of professionalized police forces relied on voluntary provisions rather than mandatory clauses (Hart, Reference Hart1955; Philips and Storch, Reference Philips and Storch1999).Footnote 6 Legislation enacted during the following years, such as the Parish Constables Acts of 1842 and 1850, even allowed counties to appoint superintending constables to supervise traditional parish constables (Emsley, Reference Emsley2008).
By contrast, the County and Borough Police Act of 1856 made it mandatory for counties and boroughs to establish a professionalized police force and became a major shift in the previous voluntary schemes (Philips and Storch, Reference Philips and Storch1999). The act established an enforcement mechanism in the form of Inspectors of Constabulary under the authority of the Home Secretary, who assessed the degree of implementation of the reform and granted up to one-quarter of the cost of provincial police forces if they were deemed efficient (i.e., if they were able to monitor their entire territory). As in the 1830s, the law was preceded by the 1853 Select Committee on Police in the Counties and Boroughs, which recommended the rationalization of provincial police forces, even though a bill to these effects failed to pass in 1854 because of the strong opposition that it raised from borough representatives (Emsley, Reference Emsley2008; Wall, Reference Wall1998).
These initial reforms were of utmost importance, as together they expanded modernized police forces to all counties and boroughs. Figure C1 in the Appendix (Supplementary material) shows that the number of modern provincial police forces went from 30 in 1835 to a peak of 233 in 1857, with spikes right after the mentioned pieces of legislation. This went along with increasing numbers of provincial policemen (from around 7,500 to 12,000 during the 1850s).
The second pattern that we observe in Figures 1 and 2 is that the highest level of police salience occurred during the 1880s and 1890s. This is somewhat surprising, given that these decades have not attracted as much attention from historians as earlier and later periods. The bottom panel in Figure 1 shows that police reform was almost as salient as fiscal policy in the late nineteenth and twentieth centuries, even surpassing it in the 1880s and early 1890s. Figure 2 confirms the importance of these decades, as discussions on the police – according to the labels assigned by Hansard – represented a large share of legislative activity, reaching up to 6 percent of all debates in the late 1880s.
The high salience of the police in the 1880s and 1890s may be related to discussions about the centralization and standardization of provincial police forces. Already in 1874, the Police Expenses Act centralized control by increasing the amount of government grants to county police forces from one-quarter to one-half, but it failed in its attempts to reduce the number of borough police forces (as intended by the Home Secretary). In 1877, the Municipal Corporations (New Charters) Act forbade the creation of a separate force for any new borough with a population of less than 20,000, while the Local Government Act of 1888 compelled small boroughs with less than 20,000 inhabitants to join their police forces with those of the adjoining county. There had been initial proposals to put them fully under the authority of the central government, but they were finally dropped from the legislation due to intense opposition (Wall, Reference Wall1998). As can be seen in Figure C1 in the Appendix (Supplementary material), the act succeeded in reducing the number of provincial police forces in England and Wales from 231 to 183 (see also Emsley, Reference Emsley1996, 93). In the process of homogenization of provincial forces, borough, and county officers adopted the formal title of chief constable in the late 1890s (a reform not welcomed by county chief constables who saw borough forces as inferior in rank).
The end of the century also sparked concerns about rising crime and public disorder, for instance, related to the 1886 mass meeting of the unemployed in Trafalgar Square or the so-called Bloody Sunday in November 1887, during which the police repressed a meeting called by the Metropolitan Radical Federation in Trafalgar Square (Emsley, Reference Emsley1996, 66–67). In 1886, parliament constituted the Disturbances Committee, which concluded that the Met lacked officers of “superior rank and education” and proposed to improve the chain of command, along the lines of the 1879 Committee on the State, Discipline and Organization of the Met (Wall Reference Wall1998, 24). In the same vein, the 1890 Police Act made it easier for chief constables in counties to ask for reinforcement from other police forces when facing social turmoil.
Likewise, these decades witnessed increased attention to working conditions and unionism in the police forces (Emsley, Reference Emsley1996). A landmark in these discussions was the 1890 Police Act, which granted pensions to police officers after 25 years of service (among other eventualities). This reform reduced the power of local authorities, as it stripped away the discretionary authority to grant pensions that had been given to chief constables (in the counties) and watch committees (in the boroughs) by the 1859 Police Pensions Act (Williams, Reference Williams2014). Political rights of police forces were also on the table. The 1887 Police Disabilities Removal Act enfranchised county police forces that qualified to vote under the franchise legislation (before this, only borough forces could vote).
The third pattern in Figures 1 and 2 is that the turn of the century initiated a period of relatively high police salience. Even though the average police salience decreased thereafter (due to the attention that parliament was paying to other issues), policing became a distinct and specialized policy issue, with many debates devoted to it. This was the time when discussions about working conditions in the police forces flourished. Prolific discussions about a weekly rest day in the late 1900s gave way to concerns about police pay amid wartime inflation during World War I, which prompted the emergence of the National Union of Police and Prison Officers and the 1918 strike by Metropolitan and City of London Police forces. This led to the 1919 Desborough Committee’s reports (1919 and 1920) and the recommendation of, among other solutions, pay rises, uniformity in pay and conditions across the country, and the establishment of a formal body to negotiate issues related to working conditions.
In sum, this first set of analyses suggests that legislative debates on the police followed three distinct stages: a first period marked by episodes of major institutional reforms (1830s to 1850s), a second period of heightened attention to the police (1880s–1890s), and a third period of constant (albeit less pronounced) police salience (1900s–1940s).
What? The content of police debates
To examine the specific issues that were discussed in parliamentary debates on the police, we ran a dynamic KeyATM on all police debates (as defined in Figure 2) from 1830 to 1945. As a reminder, KeyATMs rely on the distribution of words across documents (debates, in our case) to estimate for each document the relative prominence of specific topics, which have been predefined by the researcher. We specify ex-ante nine topics based on the most common words in the titles of debates on police issues, as well as on the historical literature on the drivers of police reform: (1) Ireland, (2) expansion of police forces to the provinces, (3) war, (4) working conditions, (5) motorization, (6) women, (7) crime, (8) public order, and (9) cost of service. In Appendix B (Supplementary material) we describe in detail the process of topic selection. We also report additional models in which we specify additional pre-defined topics (namely, Scotland, India, and Metropolitan Police). Those models yield very similar results to the baseline model presented below.
Figure 3 plots the relative topic proportion of the nine predefined topics in our corpus of police debates. The plots suggest a periodization of the topics shaping legislative debates on the police consistent with the patterns presented in the previous section. First, the topic “Provinces” is highly prominent from the 1830s to the 1850s, a pattern congruent with the emphasis that the previous literature has put on the expansion of police forces to the counties and boroughs during this period. The figure shows a slight uptick in the late 1880s when discussions about the role of the central government in police issues remerged with the amalgamation of local police.
Figure 3. Topics discussed in police debates (Key-ATM).
Notes: Topic proportions for the topics with pre-selected keywords based on a dynamic KeyATM.
Second, during the 1880s and 1890s, discussions related to Ireland became particularly prominent. This is a remarkable finding since the topic attained a level of attention in parliament only matched by that of the provincial police in the early decades of the period. As noted above, discussions around reforms of local police forces in the counties were revived during those years, while other topics also began to capture legislative attention at the time, such as those related to working conditions. Nevertheless, the model indicates that Ireland, as a distinct topic of discussion, was far and above more prominent in police debates than these other policy issues. To this, we will turn in the following sections.
Finally, the topics related to war and working conditions, as captured by the topic on salaries, became more prominent during the first decades of the twentieth century. The World Wars put pressure on police forces as they assumed new tasks, such as the protection of strategic places or the search for deserters while being depleted of men sent to the front (Emsley, Reference Emsley2008). Moreover, World War II sparked efforts to centralize command under civil commissioners and the amalgamation of several police forces. Issues related to police pay also occupied members of parliament (as in the 1919 Desborough Committee) and local authorities responsible for police forces.
Additionally, the early twentieth century witnessed the emergence of new topics, such as the importance of motorization. Technological change, especially the mass production of motor vehicles, slowly shaped the way the police interacted with citizens. The 1930 Road Traffic Act generalized the use of motor patrols across police forces. Similarly, the role of women in the police also drew attention. The establishment of the Voluntary Women Patrols in 1914 and the hiring of forty women as full-time members of the Met in 1916 initiated a path of formal involvement for policewomen, taking over some of the tasks that had been carried out by police matrons since the late nineteenth century. World War II contributed to lessening the resistance by chief constables to employ policewomen, even though their duties were still confined to traditional gender roles (e.g., the pursuit of women offenders and the care of women and children victims of crime) (Emsley, Reference Emsley2008; Lock, Reference Lock2014).
Surprisingly, crime does not feature very prominently, probably because it was obvious to all that police reform was ultimately related to these matters (albeit police work went well beyond that; see, for instance, Churchill Reference Churchill2018, Emsley, Reference Emsley1996). Public order peaks during the 1860s, in the context of debates related to the Fenian bombing campaigns (1867–68 and 1881–85), which led to the establishment of the Metropolitan Police Special Branch, and the demonstrations in favor of parliamentary reform in 1866 (Emsley, Reference Emsley2008). We present in Appendix D (Supplementary material) evidence that crime and public order salience were to a certain extent related to the evolution of crime statistics and protests, at least during the twentieth century (when we have comprehensive annual data).
The most notable finding in this section is the prominence of the Irish theme during the last quarter of the nineteenth century (reaching almost 40 percent of the topic proportion in the 1880s). This suggests that the unprecedented police salience at the time was not primarily driven by fine-tuning police reforms but was very much dominated by discussions around Ireland.
Who? Party politics in police debates
In this section, we look at the parties that participated in the debates on policing. We have classified MPs party affiliations into seven categories: Conservative, Liberal, Labour, Irish nationalist, Irish unionist, Liberal unionist, and others (see Appendix E, Supplementary material). Figure 4 shows the annual number of speeches in police debates (as defined in Figure 2) by each party family from 1835 (the first year with party data) to 1945, excluding the residual group “Other.” In Figure E1 (Supplementary material) we present the number of interventions assigned to the group “Other” as well as the number of interventions with missing information on party affiliation.
Figure 4. Importance of political parties in police debates.
Different parties dominated debates during each of the three periods of intense legislative activity on the police. First, Liberal MPs dominated the debates on the early police reforms up until the 1860s, as we would expect considering that they led most of the reforms that took place at that time. Industrialists, whose interests tended to be represented by the Liberal Party, were, in fact, the most eager proponents of a professional police (albeit not a centralized one). Industrialists supported police reforms not only because they sought better protection of their property, but also because they recognized that the army was ill-equipped to deal with social turmoil and that exacerbated class conflict (Silver, Reference Silver and Elmsley2011).
Second, the twentieth century, particularly after World War I, was characterized by the dominance of Labour and Conservative MPs and the gradual decline of Liberals, a pattern consistent with the general readjustment of party politics in the United Kingdom at that time.
Third, and most interestingly to our purposes, the 1880s and 1890s witnessed an abrupt increase in the number of interventions by Irish nationalists and, to a lesser extent, Irish and Liberal unionists. Their prominence receded already in the early twentieth century, although it did not fade away completely until Irish independence. This result aligns well with our previous findings, as this was the period when the topic “Ireland” in our KeyATM models gained importance.
Further evidence links more decisively the prominence of Irish parties with discussions about police in Ireland. Figure 5 presents the results of a covariate KeyATM with political groups instead of decades as a covariate. This allows us to examine the extent to which parties promoted different topics in Parliament. Results indicate that Irish nationalists (and to a lesser extent Irish Unionists and Liberal Unionists) were the most vocal parties on Irish matters (and these were in fact their main issue of attention). The other party families did not experience this level of disparity between topics.
Figure 5. Topics discussed by political parties (Key-ATM).
Notes: Topic proportion of the topics with pre-selected keywords by party in a covariate KeyATM.
How? Parliamentary conflict around police reform
We now turn to examine the intensity of legislative conflict in debates related to policing. First, we use sentiment analysis to estimate variation in the tone of the debates about police reform (as defined in Figure 2). Specifically, we calculate the standard deviation of the sentiment score of all interventions in a debate, assuming that higher variation reflects more rhetorical conflict. The left panel in Figure 6 shows our indicator of conflict in each debate on the police (higher values in the y-axis indicate higher levels of conflict in that debate), while the right panel shows the average levels of discursive conflict by year. Confrontation in parliament seems to increase precisely in the 1880s and 1890s (reaching the peak of the annual averages) and, to a lesser extent, during the inter-war period. Thus, the late nineteenth century was not only a period with high police salience, but it also appears to be particularly contentious.
Figure 6. Rhetorical conflict in parliament.
Notes: The left plot represents the standard deviation of the sentiment scores by debate, whereas the right plot represents the yearly average of the standard deviation of the sentiment scores by debates.
In Figure 7, we present two alternative measures of legislative conflict based on roll-call vote data. The left panel shows the number of divisions in Parliament related to the police. Since not all bills end up with a vote if there is consensus, higher numbers of divisions indicate not only greater parliamentary activity but also higher levels of disagreement among MPs. The figure indicates that during the 1880s there was, in fact, an unusual number of divisions related to policing. The right panel displays the percentage of positive votes in each division, showing that during this same period, the percentage of positive votes tended to be closer to 50 percent, which signals greater disagreement between MPs (by contrast, percentages close to 100 percent or to 0 percent signal more agreement in favor or against a given motion or bill). All these measures point towards particularly contentious debates in parliament at that time.Footnote 7
Figure 7. Divisions and votes in parliament related to police issues.
Notes: The left panel plots the number of divisions on police matters (based on the titles of the divisions) in the House of Commons by year. The right panel presents the percentage of “yes” votes in the same sample of divisions. Source: Eggers and Spirling (Reference Eggers and Spirling2014).
To summarize, our analyses of the timing (when), content (what), actors (who), and contestation (how) of police debates is consistent with established knowledge about the legislative politics of police reform. However, they also helped us discover a period of frequent legislative discussions around policing in the 1880s and 1890s, which largely focused on Irish affairs, were dominated by Irish MPs, and showed high levels of contestation. Even though those debates do not seem to have been associated with major institutional reforms (and thus have remained under-explored by the historical literature), they deserve further scrutiny given their prominence in the legislative agenda at the time.
Parliamentary debates about the police in Ireland
In this final section, we examine in detail the legislative debates around police issues in Ireland between the 1880s and 1900s. We first identify all debates that focused on Irish matters by selecting debates that include “Ireland” or “Irish” in their titles. As a second step, we keep only debates that mention at least one of our police keywords in the text of the speeches. Therefore, these debates were not necessarily about police reform in Ireland, but they were likely to be related to policing. In Appendix G (Supplementary material), we present an alternative approach in which we first select police debates (as in Figure 2, based on police keywords in the titles) and then keep a sub-sample by identifying debates with the keywords “Ireland” or “Irish” in their titles. Results are very similar between these two approaches, but the former gives us more confidence that we are not missing relevant debates specific to Irish affairs.
Figure 8 displays the number of selected debates. Police issues in Ireland seem to emerge as a topic of discussion on several occasions during the first half of the nineteenth century, most notably in the 1830s, around the time when the 1836 Irish Constabulary Act merged the Peace Preservation Force (created in 1814 by the Chief Secretary for Ireland, Robert Peel) with the County Constabulary (created in 1822 to supplement Peel’s force). We also see peaks in the mid-1840s, when police training was centralized in Dublin (ending the tradition of local training) and the central government became responsible for police payment (Conway, Reference Conway2014).
Figure 8. Debates about police issues in Ireland.
Notes: The dark line shows the number of debates on Ireland (defined as debates with the keywords “Ireland” or “Irish” in their titles) that feature at least one of our police keywords in the text of the speeches. The red line restricts the previous sample to debates with over 1,000 words.
Nevertheless, these episodes pale in comparison to the spike observed in the 1880s and, to a lesser extent, the 1890s, very much in line with the results of our previous sections. A qualitative analysis of these debates shows that many of them were short discussions between Irish nationalist MPs and the government over specific incidents and, in some cases, the nature of the police and the purpose they served. Even though the period witnessed the enactment of relevant pieces of legislation that entailed long discussions, such as the 1881 Protection of Persons and Property (Ireland) Act, the 1882 Prevention of Crime (Ireland) Act, or the 1887 Criminal Law and Procedure (Ireland) Act, these debates do not seem to explain the unparalleled rise in parliamentary attention observed during these years. This can be seen by looking at the red line in Figure 7, which features only long debates with over 1,000 words. Unlike the 1830s and 1840s, when debates on the police tended to be long, the exceptional number of debates on the police in the 1880s was related to shorter interventions.
As mentioned before, the Irish Constabulary, unlike the English police, was a centralized force with a strong military influence (Ansell and Lindvall, Reference Ansell and Lindvall2020; Boyle, Reference Boyle1972; Emsley, Reference Emsley1996), to the extent that most senior positions were initially filled with army officers and generals (Conway, Reference Conway2014). Our corpus of debates suggests that Irish nationalist MPs were using the parliamentary arena to complain about police misconduct, the military nature of the police, and the lack of control over police matters in general. As an illustration of these short but scathing debates, on May 16, 1881, the Irish nationalist MP Timothy Michael Healy, a notorious member of the Irish Parliamentary Party and, later, the anti-Parnellite Irish National Federation, complained to the Chief Secretary of Ireland that the police had torn down a placard in Enniscorthy for summoning citizens to protest the collection of rent by the sheriff. In doing so, Healy asked “whether he approved the conduct of the constables in pulling down the placard, and by what law or authority it was done,” to further notice that “force seemed now to be the only rule in Ireland” and that the “placard incident was a sample of what was occurring every day in the country.” He therefore expected “that ten times more outrages would occur, those outrages being the only resource which those unfortunate people had against highhanded proceedings” (Hansard, 1881, column 549).
Another telling example can be found on May 22, 1884, in a debate initially about the need to appoint an extra police force in the city of Cork. Several Irish nationalist MPs used the opportunity to complain about the role and fiscal burden of police forces in their own cities and constituencies. For instance, in a passionate intervention demanding the removal of the extra police in the city of Limerick, the Irish MP William O’Brien, founder of the 1898 United Irish League, stated that “it was perfectly well known that the police were a purely military force, and utterly useless for municipal purposes. … Although the city was one of the quietest, the police had done their worst to torment the people; they assaulted them in the streets; and it was not more than two years ago that they had shot them down with buckshot” (Hansard, 1884b, column 1122). On these grounds, he assured that “the Irish people were perfectly willing to pay their police as the English people did, if Parliament would give them some control over the police. But Parliament dared not give the people that control, but paid the police themselves, in order to keep the people in subjection, and then were surprised if the people did not submit tamely” (Hansard, 1884b, column 1134).
The absence of mechanisms of accountability for Irish forces (which contrasted with the situation in England, where an Inspectorate of Constabulary was already established in 1856 under the County and Borough Police Act) meant that not only police interventions in political affairs and disturbances were controversial, but also common crime control could create distrust among the population (Conway, Reference Conway2014). This clearly emerges in some of the debates related to the application of the Prevention of Crime (Ireland) Act enacted in 1882, as Irish nationalist MPs reported alleged police misconduct to the government through Parliament. For instance, the Irish MP Arthur O’Connor complained on August 12, 1884, that a sergeant of police broke into a private residence at night behaving in an offensive manner (Hansard, 1884a, column 646). Similarly, MP James O’Kelly raised a question to the Chief Secretary on August 13, 1883, regarding the arrest of an individual of his constituency accused of a murder for which he had been already trialed and acquitted (Hansard, 1883, column 267).
While this militarized and unaccountable nature of the Irish police (at least from the perspective of Irish nationalists) had been present since its introduction in the early nineteenth century, the active involvement of Irish MPs in parliamentary discussions on police issues during the 1880s and 1890s was probably driven by the social and political context of the time. On the one hand, this was a period of heightened agrarian conflict, with its peak during the Land War (1879–82) and later the Plan of Campaign (1886–1891). Ireland was not new to political turmoil, as agrarian uprisings had been commonplace in previous decades. However, the Land War, which originated in the context of the severe 1877–79 economic crisis and a growing intolerance to rural evictions and tenure conditions, witnessed particularly high levels of political violence (Garvin, Reference Garvin1983). Figure 9 shows the high level of agrarian outrages during this period (around half of them related to landlord-tenant relations and against tenants taking evicted holdings), coinciding with increasing rates of homicidal violence and indictable assaults (Finnane, Reference Finnane1997). In fact, the near-famine conditions generated by the economic crisis of the late 1870s and the ensuing political conflict resulted in the highest levels of excess mortality rates in Ireland for the entire period 1864–2022 (McLaughlin and Whelehan, Reference McLaughlin and Whelehan2024).
It was in this context of economic crisis and widespread protests that the Irish National Land League was created, a nationalist and agrarian movement built upon the agreement between the Fenians and the Home Rule League’s MP Charles Stewart Parnell to advance the cause of agrarian reform and legislative autonomy for Ireland (Jackson, Reference Jackson2003). Renamed as the Irish National League in 1882, and with a stronger focus on legislative independence, the league would become, according to Garvin (Reference Garvin1983), the first mass political party in Ireland since the O’Connell’s Repeal Association. Under firm control by Parnell, the Irish National League, and the Irish Parliamentary Party won 85 of the 89 seats contested in Ireland in the 1885 elections, the first ones after the 1884–85 electoral reforms that reduced the disproportionate political power of wealthy Protestants and strengthened the Home Rule cause (Jackson, Reference Jackson2003). Importantly, these 85 seats equaled the Liberal majority over Conservatives in the chamber, giving the Irish Nationalists political leverage. This political context gave Irish nationalist MPs an ideal platform to bring to the House of Commons their grievances and demands.Footnote 8
Parnell’s dismissal in 1890 initiated a period of party fragmentation (most notably with the split of the anti-Parnellite Irish National Federation, but also between the anti-Parnell members themselves) that lasted until the creation of the United Irish League in 1898 by the anti-Parnellite William O’Brien, and the subsequent reunification of Irish nationalist factions in 1900 under the leadership of MP John Edward Redmond (Jackson, Reference Jackson2003; Mulvagh, Reference Mulvagh and Bartlett2018). Despite the tensions that characterized this period, the party enjoyed again remarkable electoral success in the counties outside of eastern Ulster (Foster, Reference Foster2001; Garvin, Reference Garvin1983).
Alongside these shifts in the electoral ascendancy of Irish nationalists, the social background of Irish MPs also changed notably: while 70 percent of them were landlords or landlords’ sons in 1868 (73 out of 105), the presence of landowners declined to less than 9 percent in the turn of the century (6 out of 87 in 1892–95 and 8 out of 89 in 1895–1900, or up to 11–16 percent if we include farmers). The most common occupations by the end of the century were barristers, journalists, and local merchants (also with a smaller presence of tenant farmers and labor leaders), which reflected more closely the social base of the nationalist movement (Reference GarvinGarvin, 1983; Lyons, Reference Lyons1951). This was the result of parliamentary reform and the ensuing broadening of the electorate, such as the Franchise Act of 1884 and the Redistribution Act of 1885, which might contribute to explaining the new political strategy of parliamentary action by the Irish nationalist movement.
While the combination of a tense social context and landslide victories of an ascending political party helps us understand the rise in parliamentary attention to police issues, the subsequent decline observed at the turn of the century might reflect the relatively peaceful relations between the RIC and the community at that time. Civil duties were progressively assigned to the Constabulary (such as weights and measures inspections or road regulation), driving them away from public order operations and making them part of Irish daily life. The composition of its personnel also began to match the Irish society more closely (e.g., with a higher share of Catholic and married constables), a process that Lowe and Malcolm (Reference Lowe and Malcolm1992) define as one of “domestication.” This coincided with a reformist agenda by the British Conservative governments (in alliance with Liberal Unionists), which included the enactment of Land Acts (1891, 1896, and 1903) that facilitated (and even incentivized) tenant purchase of land, as well as the Local Government (Ireland) Act of 1898, which democratized local governance with elected county and district councils (Dooley, Reference Dooley2004). Liberal governments, on their part, introduced no less than three Home Rule bills in 1886, 1893, and 1912, the first two defeated by Congress and Lords, respectively, and the latter suspended in 1914 for World War I’s duration (Bartlett, Reference Bartlett2010).
This process of relative détente was interrupted by the unionist armed resistance to the 1912 Home Rule Bill and, most notably, by the uprising of 1916 and the social confrontation that led to the Irish War of Independence (Bartlett, Reference Bartlett2010; Conway, Reference Conway2014; Finnane, Reference Finnane1997; McGarry, Reference McGarry and Bartlett2018). It was in this context of police response to increasing social tensions that legislative attention to the Irish police spiked again (see Figure 8), with complaints against police misconduct being voiced once more in Parliament. For instance, the Irish MP John Dillon, leader of the Irish Parliamentary Party, aimed to delay the second reading of the 1916 Constabulary and Police (Ireland) Bill (related to police pay) until an amendment on alleged police indiscipline in Dublin was addressed. Dillon complained that “a vast deal of the trouble that has happened in Ireland in the past, and a vast deal of the bad feeling which at times has existed, disastrous bad feeling, between the police and the people, was due to some of these very officers upon whom the hon. and gallant Member [referring to the Ulster Irish MP James Craig] has pronounced such a tremendous eulogy” (Hansard, 1916, column 107).
A similar conflict unveiled during the discussion of the 1919 Constabulary and Police (Ireland) Bill (again about police pay, but also including issues related to police unionization and misdemeanors against the police). In October that year (already after the proclamation of the Dáil Éireann by Sinn Féin MPs), the Chief Secretary of Ireland, James Macpherson, accused Irish MP Jeremiah MacVeagh of obstructing the passage of the bill, and argued that “no police force in the world has had to encounter to the same extent revolution and disorder, and no force has shown greater loyalty, grit, and bravery,” adding that “I hope and believe this measure will strengthen their faith in the desire of the Government to stand by, them in their task, which they perform unflinchingly” (Hansard, 1919, column 403). MacVeagh, on his part, denied his role in obstructing the bill and claimed that the bill was “offered as a bribe to the Irish police. The police form part of the Army of Occupation, and the right hon. Gentleman desires to keep them snug and content: therefore he offers to increase their salaries,” emphasizing that the real object of the bill was to prevent Irish policemen from joining trade unions (Hansard, 1919, column 408). These remarks merited the reprimand from other MPs, such as leader of the Ulster Unionist Party, Irish MP Edward Carson, who argued that “the police in Ireland need no bribe to induce them to carry out their duties. They are men who at all times of the day and all times of the night are carrying their lives in their hands, and they have never made the slightest demur, so far as I know, to the onerous duties put upon them. To suggest a bribe in relation to those services is an insult to the most loyal force in the United Kingdom” (Hansard, 1919, column 412).
All in all, the debates on police issues in Ireland since the 1880s reflect the contentious nature of policing in a deeply divided society. In such conditions, where societies are defined by severe communal cleavages, the police tend to fall into what Weitzer (Reference Weitzer1995) defines as a divided society model of policing, characterized by systematic bias in law enforcement and polarized communal relations with the police. As Garvin (Reference Garvin1983) notices, Irish nationalist MPs won elections in Irish districts by wide margins but without holding governmental office, which left them with few options but airing their concerns and grievances in the parliamentary arena, supporting other party policies (as it was the case with the Liberal Party and the various Home Rule initiatives of the late nineteenth and early twentieth centuries), or implementing a strategy of parliamentary disruption (O’Farrell, Reference O’Farrell1975). This is what the recurrent complaints expressed by nationalist Irish MPs in parliament seem to reflect, as they draw attention to what they perceived as police bias against their community and refer the matter to the competent authority (the British government) that could effectively implement policy changes. Indeed, the problems associated with centralized authority and lack of local governance were acknowledged by Conservative British MP John Newman, when in 1916 complained that
In an ordinary English shire or town police officers have got to look to the local authority. If they want an increase of pay, a war bonus or what not, they have got to go to the Watch Committee or some local authority. But in Ireland the Dublin Metropolitan Police and the Royal Irish Constabulary have to look to the Treasury, and to this House, for any increase of pay or war bonus. Therefore it is necessary that matters of this sort must be discussed in this House (Hansard 1916, column 115).
In that regard, these events are quite unique to the British case, in which some degree of political representation was granted to (some) colonial territories, but local autonomy remained strongly curtailed and the coercive arms of the state resembled in some respects more an occupying army than a civilian police force.
Conclusion
Over the course of the nineteenth century, efforts to centralize power by national political elites in Western Europe forced a renegotiation of the relationship between local communities and the central government. The Industrial Revolution triggered major demographic changes, accelerating urbanization and with it a new concern for the perceived rise in criminality. It also upended economic life, breeding new political movements that spearheaded political change through popular agitation. Driven by these pressures, states increasingly adopted new tasks and began to provide a broader array of public services. Most importantly, Western European states experienced a process of growing bureaucratization, which professionalized the roles and attributions of public officials and routinized and standardized their interactions with the population (Thornhill, Reference Thornhill2008). At the same time, many of these states were in the midst of an imperial race that pushed them to project their power outwards and to exercise differentiated coercive practices against their colonial subjects (Mann, Reference Mann2012, 408–411). These processes involved a dramatic reconfiguration of the state. Changes in the purpose, instruments, and conditions for the deployment of state coercion were at the center of those political and social transformations. It is from that context that modern police forces emerged (Ansell and Lindvall, Reference Ansell and Lindvall2020, 63).
Through the analysis of a large corpus of parliamentary debates in the United Kingdom from 1803 to 1945, we have examined how legislative activity on policing related to broader social and political processes of state modernization. Our analyses confirm previous accounts of the periods of major institutional reform, mostly from 1830s to 1850s, and improvements in working conditions for police officers, particularly since the 1900s (e.g., Bayley, Reference Bayley1990; Emsley, Reference Emsley1996; Finnane, Reference Finnane, Knepper and Johansen2016; Reiner, Reference Reiner2000; Wall, Reference Wall1998). However, a focus on institutional changes marked by the passing of major legislation inevitably overlooks periods in which policing became highly salient in legislative debates but nonetheless did not culminate with concrete legislation. Our approach using computational text analysis has sought to overcome these blind spots in the literature, allowing us to identify other moments in which policing was extensively discussed and contested. In particular, our analyses show that the parliamentary debates on policing in the 1880s and 1890s underscored the tensions faced by the British state as it tried to modernize itself domestically while maintaining imperial practices in its colonial territories. The growing standardization of the exercise of coercion clashed with the persistence of differentiated arrangements to govern colonial territories. Our results highlight the ways in which actors at the time noticed these tensions and resorted to the institutional channels available to them to denounce those inequalities.
By adopting a long-term view of legislative discussions on policing, we can relate those debates to broader processes of social and political transformation. The modernization of the British state meant a reconfiguration of its coercive apparatus during the long nineteenth century, as it would increasingly turn inwards, towards disciplining and regulating its own population. This article has offered new insights into how those processes of bureaucratization and standardization in the exercise of coercion unfolded within the British Parliament, how they generated tensions with other practices perpetuated by the imperial aspects of the British state, and how different political actors at times contributed, at times contested, and at times strategically exploited the institutional changes brought about by the reconfiguration of the British state.
Supplementary material
To view supplementary material for this article, please visit https://doi.org/10.1017/ssh.2025.30
Acknowledgements
The authors acknowledge funding from the Swedish Research Council grant no. 2021-02907, as part of the project “The Politics of State Building: Studying Investments in State Capacity through Legislative Debates”. Oriol Sabaté has also received funding from the Beatriu de Pinós Fellowship Programme (2019 BP 00039).
Open access
