The law & economics of private prosecutions in industrial revolution England

Abstract

Can the market provide law enforcement? This paper addresses this question by analyzing an historical case study: the system of private prosecutions that prevailed in England prior to the introduction of the police. I examine why this system came under strain during the Industrial Revolution, and how private clubs emerged to internalize the externalities that caused the private system to generate too little deterrence. The historical evidence suggests that these private order institutions were partially successful in ameliorating the problem of crime in a period when public choice considerations precluded the introduction of a professional police force.

1 Introduction

The provision of law enforcement is traditionally held to be a public good that has to be provided by the state (for a recent statement, see Polinsky and Shavell 2007). But this view has been challenged by research demonstrating that the private provision of policing is both theoretically possible and empirically feasible (see Becker 1974; Landes and Posner 1975; Friedman 1984, and numerous others, many of whom are cited below). A number of issues, nonetheless, remain unresolved. In particular, while it is established that private law enforcement institutions can function well in small-scale and close-knit communities, it is not clear whether they are scaleable, that is, capable of generating an effective level of deterrence in larger societies.

In this paper, I examine an historical case study: England during the Industrial Revolution (1750–1850) to shed light on this question. I argue that the pre-modern private system of prosecutions broke down during the Industrial Revolution. However, there was no publicly funded, professional, police force in London until 1828 and parts of the country remained unpoliced until 1856. I show that a private-order institution was able to respond to the breakdown in law and order, and to ameliorate the problem of crime in the period prior to the introduction of the police. Clubs, known as associations for the prosecution of felons, subsidized prosecutions for their members and increased the overall level of deterrence.

The example of private law enforcement in eighteenth and nineteenth century England is important because, while most other examples of private law enforcement are based on small-scale societies, which arguably do little to inform our understanding of how such institutions would function if scaled up, England in the late eighteenth century was a large, rapidly urbanizing economy, experiencing the onset of sustained economic growth.

Economists in the Pigovian tradition identified instances of justified government provision by identifying possible market failures (Pigou 1952). This provided a clear rationale for the growth of government in the nineteenth and twentieth centuries: the rise of the state could be interpreted as representing the correction of longstanding market failures.¹ This view was entirely in keeping with the views of mid-twentieth century historians of English policing, who portrayed the traditional private system in the worst possible light: a disordered and arbitrary system that failed to provide security (Armitage 1932; Reith 1936, 1943, 1956; Critchley 1967; Tobias 1979). In their opinion, the rationale for a professional police force was clear-cut: it corrected a market failure that persisted only because of the irrational attitudes of reactionary elites.²

Revisionist historians undermined and discredited these traditional police histories in the 1970s (Silver 1967; Storch 1975, 1976; Philips 1977; Hay and Snyder 1989). Similarly, the Pigovian framework was undermined by the law and economics and public choice revolutions. The latter demonstrated that the motives of policy makers, in asserting state control over parts of the economy that had previously been privately organized, were not necessarily benign (Buchanan and Tullock 1962; Tullock 1965, 1971). The former showed that, even in the presence of externalities, government intervention may not be required in order to obtain efficiency or to provide public goods (Coase 1960; Demsetz 1970).³

More recently, economists working in the ‘economics of anarchy’ have demonstrated that private individuals are capable of organizing governance without government in a wide range of settings. Becker (1974), Landes and Posner (1975) and Friedman (1984) outlined the theoretical case for the efficiency of a private system of law enforcement and a large number of studies provided empirical evidence for the private provision of law or contract enforcement services (see, inter alia, Friedman 1979; Landa 1981; Benson 1989; Greif 1989; Bernstein 1992).⁴ A principal finding from this work and from the extensive literature in anthropology, economics, and sociology is that, since restitution and ostracism are commonly and successfully used to resolve disputes, public forms of legal enforcement are unnecessary in small-scale societies.⁵

Importantly, private systems of law enforcement rely either on repeated interaction between individuals or on reputation. In small-scale, closely knit societies, these mechanisms are powerful in enforcing cooperation: the shadow of the future induces individuals to refrain from violence or theft (see Kandori 1992; Dixit 2004).⁶

An important question that this raises is: can private forms of law enforcement be used in modern societies comprising not dozens or hundreds, but hundreds of thousands, or millions of individuals, who do not know one another and will not necessarily interact with each other in the future? There are strong theoretical reasons for thinking that, on their own, mechanisms dependent on the Folk Theorem may be unable to support cooperation in large-scale societies. In particular, if punishing offenders is costly, then private systems of law enforcement may face a free-rider problem, and this free-rider problem may become exacerbated as the society becomes larger.⁷ In sum, the literature on private forms of law enforcement has established the effectiveness of private systems of law enforcement in small-scale, tightly knit, societies, but it has not yet demonstrated that these institutions can effectively provide similar services in larger, or more fluid, societies. This is the ‘scaleability problem’.⁸

The system of private prosecutions that operated in eighteenth and early nineteenth century England allows us to address this issue. During the Industrial Revolution, a private system of prosecutions that had worked effectively for centuries came under severe strain. However, it was not replaced by a professional police force. Parliament rejected the idea of a nationwide police force in the 1780s, and it was not until 1856 that every county in England had a police officer. The unique feature of the English example, therefore, is that it allows us to study how a private system of law enforcement responded to the increase in crime that accompanied the industrial revolution.

This is not the first paper to study how the English system of private prosecutions functioned. Friedman (1995) also considers the legal system of eighteenth century England from a law and economics perspective. He discusses how private prosecution associations increased deterrence by (1) pre-committing potential victims of crime to undertake prosecutions and (2) internalizing the benefits of deterrence among members. Allen and Barzel (2011) consider how greater standardization of production shaped the evolution of criminal law, the introduction of the police, and the introduction of factory colonies in this period.

In a companion paper, I examine how the internal organization of the prosecution associations enabled them to provide public goods privately by adopting strategies of bundling and price discrimination (Koyama 2012a).⁹ This paper differs in that it focuses on how these associations emerged in response to the Industrial Revolution, which disrupted a largely private system of law enforcement that worked relatively well in seventeenth and early eighteenth century England. It explains why the Pigovian solution of a professional police force was opposed for public choice reasons and it assesses the comparative effectiveness of prosecution associations as a private solution to the problem of crime in Industrial Revolution England.

2 The private system of prosecutions

Law enforcement in pre-industrial England was largely private.¹⁰ However, England in the eighteenth century did not rely upon purely informal systems of restitution which are typical of small-scale societies. The legal system was centralized and publicly provided (although costs were defrayed through fees).¹¹ Nevertheless, the extent to which this system relied on private individuals is striking. Local magistrates, Justices of the Peace (JPs) and constables, involved themselves in assisting victims in pursuing and apprehending criminals, but, with the exception of crimes against the state, such as treason and coining of money, the costs associated with assembling a prosecution were borne by the victim (Hay and Snyder 1989, 26).¹² This system evolved little and in the eighteenth century most of ‘the country was policed by arrangements which had been inherited from the Middle Ages’ (Tobias 1979, 25).

Although there are no nationwide estimates of offenses committed, the private system of law enforcement appears to have functioned adequately in pre-industrial England. McMullan summarizes the historical consensus as follows: ‘[t]he “system of amateur policing,” drawn as it was from the local village or ward, seems in the hands of the new social history to have been remarkably able, conscientious, efficient, and adequate to the needs of rural and small-town England’ (McMullan 1987, 254). Perhaps the most compelling evidence we have for this is that the absolute number of executions fell steadily during the seventeenth and much of the eighteenth century, despite population growth and despite the fact that the number of crimes liable for the death penalty rose dramatically.¹³

A number of factors were at work limiting crime in pre-industrial England. First, the small-scale nature of the rural communities that made up most of England (outside London) meant that opportunities for property crime were rare. Valuable possessions were locked away and guarded carefully; they often were personal items that could be easily identified (and hence recovered) by their owners (Allen and Barzel 2011). Second, the existence of well-established social norms meant that a variety of alternative forms of punishment were available that could be used in conjunction with the formal legal system. Formally ‘indictable’ behavior could be deterred or punished by ‘dismissal or chastisement by an employer; informal coercion or admonition by a priest or landowner; arbitration; and control through the poor law. Ostracism and eventual ejection from the community (an informal secular equivalent of excommunication) and control through the ritualized satire of rough music or stang riding might be added to this formidable list of methods that might be brought against the deviant or the delinquent without using the formal machinery of the law’ (Sharpe 1980, 117–118).¹⁴ Third, these very same social norms also meant that the criminal justice system, when it was used, functioned effectively. Volunteer constables and Justices of the Peace assisted victims and pursued criminals because they were incentivized to do so by local social norms.

However, this changed during the eighteenth century and, by the time of the Industrial Revolution, contemporaries believed that a crime wave was underway. The view at the time was ‘that ‘crime has much increased’ everywhere in the years before Victoria came to the throne’ (Jones 1992, 51–52). No estimates of the actual number of offenses were kept until 1857, so the main evidence that we have for this crime wave comes from the increase in pre-trial committals. From 1805 onwards the number of suspects apprehended by private individuals, thief-takers, or constables, brought before magistrates, and committed to prison to await trial was recorded at a national level. These data are depicted in Fig. 1. Inevitably, it reflects attitudes towards crime and resources devoted to eradicating crime as well as the actual number of offenses.¹⁵ Nevertheless, local studies certainly suggest that property crimes were increasing in the late eighteenth century (Beattie 1986, 215).¹⁶ In order to determine why the private system of prosecutions came under strain during the Industrial Revolution, it is useful to consider what economic theory says about the incentive to commit crime.

Fig. 1

The increasing number of committals between 1807 and 1838 in absolute numbers and per 100,000 people. Source: First Report from the Commissioners Appointed to Inquire as to the Best Means of Establishing an Efficient Constabulary Force in the Counties of England and Wales (1839)

3 The problem of crime in 18th and 19th century England

3.1 The market for crime

I use the market for crime framework developed by Becker (1968) to analyze how a system of private prosecutions might work in a society like Industrial Revolution England. The supply of criminals and the demand for security of property determine the equilibrium amount of criminal activity. Potential criminals can choose between ordinary employment or criminal activities. Thus the supply of crime is a function of the cut-off level of criminal ability at which a potential criminal is indifferent between committing an offense and not committing one. This is an increasing function of the value of the goods they can acquire through theft, and a decreasing function of the probability of apprehension and prosecution and of the cost of being convicted.

Under a private system, a victim of a crime can choose whether or not to prosecute a suspect. A victim will prosecute only if the benefits exceed the costs. In equilibrium, the market level of crime will be determined by both the supply of criminals and the implied demand for crime as determined by the willingness of private individuals to prosecute.

If individuals internalize the effect that decisions to prosecute have on the overall level of deterrence, and hence on the incentive that the marginal criminal has to become a criminal, then a private and decentralized system of prosecutions can achieve the socially optimal level of prosecution and crime. What incentive did individuals face in pre-industrial England? The main direct benefit an individual obtained was the recovery of goods that had been stolen. But a prosecutor also benefited indirectly to the extent that he lowered the probability of being the victim of a crime in the future by increasing the level of deterrence (or by incarcerating a particular criminal); in a small society this probability was positive. An individual could also obtain a reputation for prosecuting all offenses against him. And, in a small-scale society, neighbors could bring social sanctions to bear on individuals who did not prosecute crimes against their own persons or property. Hence, in a tightly knit society, with strong social norms, it is possible for a private system of prosecutions to approach the efficient level of prosecutions. However, the theory predicts that as societies become larger and more fluid, the incentive individuals have to prosecute will diminish and the gap widened between the socially efficient level of prosecutions and the actual level of prosecutions.

3.2 The industrial revolution and the market for crime

The Industrial Revolution exacerbated this problem in several ways. On the supply-side, urbanization and industrialization strengthened the incentives individuals had to commit crimes—particularly property crimes. Contemporary writers argued that economic growth reinforced the ‘temptation’ to steal (Fielding 1751, 4; Colquhoun 1796, 34).¹⁷ The consumer revolution of the eighteenth century meant that there were many more valuable personal possessions that a thief could steal than had previously been the case: wrist-watches and ready-to-wear garments became common (see Brewer and Porter 1993; Lemire 1991; de Vries 2008; Styles 2007; Koyama 2012b).¹⁸

Allen and Barzel (2011) argue that theft became easier as standardized products increasingly replaced personalized items. Greater standardization lowered the probability of detection. But it was not only standardized manufactured goods that became easier to steal. Better toll roads encouraged horse theft (Styles 1989, 20). The rise of new industrial towns made faceless crimes much more common and made it easier for criminals to escape with their ill-gotten gains (Critchley 1967, 21).¹⁹

The Industrial Revolution was characterized by rapid population growth, urbanization, and structural change as much as by per capita income growth (see Crafts 1985; Harley 1999). Total output increased, but real wages were stagnant, while income inequality grew dramatically (Allen 2009).²⁰ Manufacturing boomed in Lancashire and the Midlands, but in other parts of the country, such as East Anglia, the traditional textile industry went into decline, causing widespread hardship and unemployment (Mokyr 1988; Hudson 1989).²¹ Economic growth, greater inequality and standardization increased the returns to crime directly, while stagnant or falling real wages reduced the value of the next best alternative to crime.

Social norms in pre-industrial society encouraged prosecutors to internalize the positive externalities that they generated. Urbanization eroded these norms. Moreover, as informal alternatives to the criminal justice system began to disappear, the costs of using the law courts increased as the law became more professional and the use of lawyers became more common.²²

Reformers from the mid-eighteenth century onwards noted that too few individuals carried prosecutions to conclusion.²³ The advocate of police reform, Patrick Colquhoun (1796, 212) stated this reasoning most clearly: ‘many persons who suffer by means of small robberies, afraid of the trouble and expense of a discovery and prosecution, submit to the loss without enquiry’ and in so doing provide ‘a great encouragement to thieves of every class’. It was not just that undertaking a prosecution was costly: rather there was an incentive to free-ride on the prosecutions of others, with the result that the actual number of prosecutions was below the socially optimal number. Henry Fielding (1751, 108) noted that ‘[r]obbery is an Offence not only against the Party robbed, but against the Public, who are therefore entitled to Prosecution; and he who prevents or stifles such the Prosecution is no longer an innocent Man but guilty of high Offence against the Public Good’.

The growing gap between the market and the socially optimal number of prosecutions explains the sense of crisis that pervades contemporary discussions of crime.²⁴ Colquhoun (1796, 246–247) argued that the problem was so grave that a public prosecutor was required. ‘Experience’ had taught him that the moral arguments in favor of prosecuting ‘powerful as they are, will neither awaken in the mind of men that species of public spirit which shall induce’ individuals ‘to become willing prosecutors under all the trying delays, added to the expense often of bringing a number of witnesses from the country, and keeping them hanging on in the court of justice, perhaps for several days together’. A public prosecutor, acting in combination with a professional police force, would not ‘only remove that aversion which prosecutors manifest on many occasions, to come forward, for the purpose of promoting the ends of public justice; but it would prevent, in a great measure, the possibility of compounding felonies or of suborning witnesses’ (ibid., 252).²⁵

3.3 A professional police force?

Advocates of police reform contrasted the largely private English system with the public French system. Charles Cottu (1820, 23), who visited England in the 1810s, observed that the English were uninterested in the causes of crime and were ‘very indifferent as to the condemnation of the accused, should those whom he has injured suffer themselves to be touched with commiseration, or overcome by indolence’.²⁶ The anonymous author of the Code D’Instruction Criminelle (thought to be James Mill) argued that the French system was superior to the English system in that it excelled in the ‘preliminary business’ of assembling a case, securing the suspect, and searching for evidence for his guilt. Whereas, in England, ‘[t]here is no public prosecutor’:

‘we trust entirely to the information of the individual whom the crime has injured. For want of a prosecutor, we bind this voluntary informer to prosecute; which is imposing upon him so great a burden, in loss of money, loss of time, and in trouble of various kinds, that, in a great proportion of instances—in, probably, by the majority of instances—where the injury is not of an atrocious sort, the injured person conceals it, and withholds complaint. Instead of taking measures to secure the notice of crimes, we thus take measures to secure their concealment’ (Anonymous 1810, 108).

From the point of Utilitarian reformers, as for modern Pigovian economists, the solution to this problem seemed obvious: introduce a publicly funded and professional police force and public prosecutor.

Parliament did discuss introducing a publicly funded police force as early as the 1780s. However, such a force was not introduced for several decades and there was no nationwide force until 1856. Opposition was widespread. Hay (1975, 18) observes that the rural ‘gentry would not tolerate even the idea of a police force,’ as ‘they remembered the pretensions of the Stuarts and the days of the Commonwealth, and they saw close at hand how the French Monarchy controlled its subjects with spies and informers’. Rural elites also controlled the magistrates and retained discretionary authority over local constables which they stood to lose if a national police force came into being.

The fears of the gentry were shared widely. A willingness to take on the responsibility of a prosecution privately was seen as the price of a liberal society, and it was believed that ‘[s]ome degree of trouble and expense in the vindication of our rights is the condition of a free government’ (Anonymous 1836, 34). The legal theorist William Paley (1830, 164) stated that

‘[t]he liberties of a free people … permit not those precautions and restraints, that inspection, security, and control, which are exercised with success in arbitrary governments … least of all will they tolerate the appearance of an army force, or of military law, or suffer the streets and public roads to be guarded and patrolled by soldiers; or lastly, entrust the police with such discretionary powers, as may make sure of the guilty, however they involve the innocent’.²⁷

Whig reformers who aimed at overturning many aspects of the old eighteenth century criminal justice system (such as the reliance of capital punishment) shared Paley’s opposition to a police force. One historian observes that ‘to the notion that the law might be more effectively enforced if there were a reformed and efficient police, they exhibited in general a notorious repugnance. If in this they were partly influenced by a dislike of putting stronger instruments of power in the hands of the government, they were also manifesting a deep-rooted popular prejudice which characterized all parties and all classes’ (Gash 1961, 320).²⁸ A public police force might not be used to crush English liberties directly, but it could still be an inefficient and bloated source of patronage for the King’s ministers, and thereby corrode these freedoms indirectly. Given that in late eighteenth and early nineteenth century administrations, there was little distinction between private and public office or the associated income streams, these fears were neither irrational nor necessarily driven by ideology.²⁹

4 Prosecution associations

The private system of prosecutions was unable to cope with crime in Industrial Revolution England. The Pigovian solution to the problem of crime in early-nineteenth century England was off the table for public choice reasons. This gives us a unique opportunity to see how private-order institutions were able to respond to a recognized market failure and to address the scaleability critique.

Beginning in the mid-eighteenth century, private associations and organizations began to be formed across the country. These associations emerged out of agreements neighbors made with one another to prosecute any crime committed against them. Over time, the agreements became more formal. Associations drew up constitutions, membership lists, and rules that committed them to fund prosecutions on behalf of their members. Members paid entry fees and annual subscriptions which were used to help pay for the expenses involved in apprehending and prosecuting a suspect. Most associations had their own solicitor who represented the association in court and provided legal advice to members. The funds of an association were also used to pay informers and to advertise rewards.³⁰

4.1 The emergence of prosecution associations

The surviving data on associations for the prosecution of felons are patchy. Nevertheless, the data that is available suggests that the number of associations grew rapidly after 1780. There are data on 534 separate associations. However, a great deal of information is missing for many of these associations. For example, unless an account book for an association survived we do not know how many members it had, what its rules were or how many cases it prosecuted; for many associations we do not even know when they were founded or when they were dissolved. The surviving account books do, however, provide a wealth of qualitative information about how associations operated. But the small number of associations for which we have this kind of detail precludes using this information in a regression framework. Figure 2 plots the founding dates for the 215 associations for which it is possible to identify a founding date. Associations began to emerge in noticeable numbers only after 1750; the number of associations grew rapidly in the last two decades of the eighteenth century and the first two decades of the nineteenth century.³¹

Fig. 2

The number of associations founded per decade. The dates refer to major demobilizations. The total sample consists of 215 associations

Figure 2 indicates that prosecution associations emerged at the same time as contemporaries argued that England was in the midst of a crime wave. Contemporaries held that crime rates spiked in the aftermath of major wars as large number of demobilized soldiers and sailors returned to civilian life (Beattie 1986, 213–235). Consistent with this view, the number of associations founded spiked at the end of the American War of Independence and at the end of the Napoleonic Wars. The expansion of prosecution associations was associated with the marked increase in the number of committals depicted in Fig. 1.

4.2 The sizes of prosecution associations

The theory of club goods explains how these associations were able to internalize the externality of deterrence among members (Buchanan 1965). Clubs provide goods to members that are non-rival (to a certain point) and excludable. Prosecution associations functioned as economic clubs. An association was committed to prosecuting all crimes against members in return for a membership fee. It generated a direct and excludable benefit for members, but it also provided an indirect and non-excludable benefit to non-members as it increased the level of deterrence and thereby reduced the total amount of crime. In order to survive, associations had to provide sufficient benefits to members to justify their membership fees.

The club goods framework predicts that an association would first attract members with valuable property to protect as they were most likely to be victims of crime and also most likely to prosecute crimes in any case. This prediction is consistent with the evidence. The membership of rural associations comprised local landowners and did not include the majority of the population.

Nevertheless, it is important to emphasize that the number of individuals who joined associations for the prosecution of felons was not insubstantial. It is impossible to gauge systematically the average size for most of the associations for which we have records, as these fluctuated year-by-year. But we can look at a snapshot of rural associations’ sizes by examining the number of members of those associations that advertised in the Ipswich Journal in January 1800. This Suffolk newspaper contains advertisements from five rural associations, the Barking Association which had 27 members, the Benacre Association with 60 members, the Bungay Tunns Association which had 34 members, the Hoxne Association with 54 members, and the Stadbrook Association which comprised 31 members (Ipswich Journal 1800, Saturday, January 4). It is likely that these associations represented a significant fraction of the property owners in these communities. For example, the membership of the Wheldrake Association, in Yorkshire, was limited to a rural parish of a few thousand acres and, when it formed in 1816, its 50 members comprised the majority of farmers in the area (Wheldrake Association 1816–1864). Urban associations were larger. When the Hammersmith Association was reformed in 1818, it initially comprised 188 members (Hammersmith Association 1818). The Association of Cawthorne, a large coalmining village outside the industrial town of Barnsley had at first 173 members in 1843 (Cawthorne Association 1843).

The club goods framework predicts that where property is highly unequally distributed as in the countryside, association membership was likely to be restricted to landowners; but where it was more evenly distributed, as in the cities, membership was more likely to be diverse. This is consistent with the evidence. While the membership of many rural associations comprised landowners, urban associations like the Cambridge Association drew on a much broader membership base that included merchants, shopkeepers, blacksmiths, brewers, bankers and textile workers (Cambridge Association 1841).³²

Maintaining membership was a critical problem for all of the associations. Associations bundled other benefits (such as annual dinners and social occasions) with membership and relied on social prestige to attract and retain members (see Koyama 2012a, 14–15). They also practiced price discrimination, which allowed them to reduce their membership fees in order to attract individuals with less property to protect (and, hence, less demand for the services provided by the association). Some associations tied membership fees to income or wealth. Others based their membership rates on property values.³³ The Caddesley Corbett association charged its poorest members just 10s and its richest members £1 (Caddesley Corbett Association 1799).³⁴

4.3 The number of prosecutions and marginal deterrence

Some historians have suggested that associations for the prosecution of felons were a peripheral phenomena that did little to reduce overall levels of crime because they carried out only a relatively small number of prosecutions (see Schubert 1981, 28–29; Beattie 1986, 49–50). In fact, it is difficult to estimate the proportion of prosecutions were undertaken by prosecution associations. Many associations were small and prosecuted only a few criminals, but some were remarkably active. The City of York Association recorded (and offered rewards relating to) 10 to 20 offenses annually and appears to have prosecuted successfully between 5 and 10 of them during the 1850s (City of York Association no date). Philips (1989) surveyed a large number of associations and he found that while on average many associations prosecuted a relatively small number of cases each year, in some periods associations could be involved in a relatively large number of them. The Harleston Association in Norfolk was involved in seven prosecutions per year between 1808 and 1815. More than 13 criminals were prosecuted in an average year by the Salt Hill Association between 1836 and 1860 (Philips 1989, 167–168).³⁵

However, the true measure of the impact of prosecution associations was not the number of prosecutions they undertook, but which crimes they prosecuted. Associations were publicly committed to prosecuting all crimes suffered by their members. However, there is evidence that they were more selective. For example, there are only two known cases of associations prosecuting murder cases (Philips 1989, 145). No matter what, victims prosecuted serious crimes against them (with the assistance of local magistrates and constables). Associations concentrated on prosecuting property crimes and the cases recorded in the account books often pertain to extremely minor crimes.³⁶ This suggests that one achievement of prosecution associations was to increase deterrence of crime at the margin. Theoretically, the reason why the existence of prosecution associations can reduce the overall level of crime significantly is that the additional prosecutions in which an association would engage are precisely those marginal cases that private individuals acting on their own would not prosecute. Hence, an association that prosecuted only a small number of cases could nevertheless make a decisive difference in terms of increasing the overall probability of prosecution.

4.4 Property crimes and public order crimes

The above analysis suggests that associations should have predominantly been concerned with property crimes. The free-rider problem to which they emerged in response was likely to be most acute for theft and robbery. Violent crimes (especially rape and murder) were likely to be prosecuted regardless of the cost-benefit calculation of the individual victim. Associations had little incentive to prosecute more general ‘public order’ offenses such as disorderly behavior or rioting.³⁷

Advocates of police reform criticized constables and associations for not prosecuting public order offenses. Associations acted against specific crimes committed against members. The Parliamentary commissioner, William Henry Newham, wanted the police to be used to remove paupers from cities and towns, and noted that a regular police could be used to suppress the ‘great evils that we suffer from … the ill-regulated state of the beer-shops, pilfering, poaching, petty offences’ (Evidence taken by the constabulary force 1839, 17).

In the wake of the French Revolution, and the Peterloo Massacre of 1819, public fear of rioting and civil disorder increased. Prosecution associations and other informal forms of policing were relatively ineffective in dealing with riots. The new police forces, first introduced in London, in contrast, were seen as particularly useful ‘in meeting the recurrent concern with arson, riot, and politicized disorder as a background motif of the first half of the nineteenth century’ (Hay and Snyder 1989, 6). Indeed, elites outside London accepted the County Police Act of 1839 in reaction to the riots and public disorder associated with Chartism (Hart 1955, 426).

There was another side to this, however. While local prosecution associations protected the property of the middle and upper classes from the depredations of the poor, the new police introduced in the 1830s were employed to reform the manners and behavior of the working classes (Storch 1975, 1976, 1989; Davies 2002).³⁸ The police punished crimes that previously had gone unpunished under the traditional system. They increasingly became involved in suppressing ‘immoral behavior’, notably prostitution, gambling and public drunkenness. Hay (1980, 58) summarizes the revisionist view that ‘the new police were frequently the agents of a middle-class assault on popular mores, not just crime or riot, and they introduced constant surveillance into working-class communities which had long since escaped the knowledge of squire and parson’. Hence, the shift from private to public policing was, in part, a response to the greater demand for protection against riots and the maintenance of public order. However, once the police were created, they were used to prosecute behavior that had not previously been considered criminal.

4.5 Specialization and integration

Another notable feature of prosecution associations was the extent to which they were vertically integrated. They were often involved in all stages of a criminal investigation from the apprehension and arrest of suspects to securing convictions in court. This high degree of vertical integration made sense. There were fixed costs involved in setting up an association and members could not be expected to be involved in numerous different organizations, each providing separate but related services. In a world of high information costs, associations harnessed the local knowledge of individual members—who were typically long-time residents, rooted in their community and in possession of specific knowledge that no outsider could easily know.³⁹ Association meetings provided forums where this information could be disseminated among individuals who had an incentive to use it to identify likely suspects, locate criminal hiding places, or get witnesses to come forward.⁴⁰

The local character of most associations, however, meant that there was no formal communication or cooperation between different associations. The professional police were less vertically integrated. They were involved in the apprehension and arrest of suspects, but did not prosecute them directly until the late nineteenth century. But they were more horizontally integrated, even though each local force retained a high degree of independence.

The downside to the absence of any horizontal integration across associations was a lack of specialization. Associations may have perfected amateur policing, but they did not provide opportunities for individuals to acquire any particular skill in investigating crime: ‘formal detection was mostly ad hoc, contingent on the good will and free time of law-enforcers who were not experts in gathering information or acting upon it’ (Shpayer-Makov 2011, 19). It was always possible to hire specialists but the small size of most associations meant that they pursued only a small number of crimes and this ensured that they typically could not afford to employ private detectives.⁴¹

After the 1856 County and Borough Act, police forces were directed to cooperate with one another. This could be a source of tension. Conley (1991) reports several clashes between county and borough police forces in the 1860s.⁴² However, it also enabled police forces to share information and skills. The 1830s and 1840s saw the expansion of the rail network across Great Britain, and the 1850s saw the introduction of the telegraph. These developments made horizontal integration increasingly valuable. From mid-century onwards police forces began to train specialist detectives who could be assigned to investigate particularly difficult cases (Shpayer-Makov 2011). In 1842, Scotland Yard—the first modern detective agency—was established within the Metropolitan police. These detectives specialized in solving cases that required a degree of expertise and training such as fraud, forgery, embezzlement, extradition cases, or matters involving foreign governments. Scotland Yard was ‘where real professionalism was expected to be found’ and these detectives were sent elsewhere in the country as police specialists: with ‘the establishment of an official detective department at Scotland Yard, and the reputation it acquired as an elite force, communities with and without police detectives applied to the Yard for aid in serious and unsolved cases’ (Shpayer-Makov 2011, 45). Thus, as the opportunities for productive specialization in detecting crimes increased, so did the relative merits of a professional police force over the older private prosecution associations.

5 Conclusion

Pre-industrial England relied on a private system of law enforcement. Victims and their families were responsible for prosecutions. This created an incentive to free-ride and under-provide deterrence. Nevertheless, the available evidence suggests that the system functioned effectively enough so long as communities remained small-scale and so long as informal norms supported it and limited the incentive victims had not to engage in costly prosecutions.

Industrialization exacerbated the free-rider problem. Critics of the private system argued that a professional police force and a public prosecute should replace it. Importantly, this did not happen for several decades, and in the meantime, individuals increasingly replied on private associations for the prosecution of felons. Associations emerged out of local and informal agreements amongst neighbors, but they ‘scaled-up’ and were sufficiently successful that they spread across the country during the late eighteenth and early nineteenth centuries.

Drawing on the theory of club goods, this paper has shown how these associations could ameliorate this free-rider problem and increase the overall number of criminal prosecutions. This analytic narrative cautions against extrapolating from institutional mechanisms that can function in small-scale societies and arguing that they can be relied upon in the large, open-access, market societies, while also suggesting that private-order legal institutions can be scaled-up to a considerable extent. However, it is important to note that this scaleability was limited. While prosecution associations partially rectified the problems associated with individual prosecutions, the limits imposed by distance and geography meant that they remained relatively small.

Prosecution associations had shortcomings. They could not provide preventative policing. They responded after the fact to crimes and struggled to undertake large-scale operations against well-organized criminals. Some associations, notably the Barnet Association, were successful in maintaining a permanent watch, but many other societies found it too costly to do so (for more details see Koyama 2012a, 117–120). Associations often formed in response to specific threats and once these threats were overcome they might then disappear. Charles Bathurst of Gloucester in answer to the Constabulary Force Commissioners, describes a successful association at Cheapstow that he suggests had formed to break up a sheep-stealing gang and was now inactive (Evidence Taken by the Constabulary Force Commissioners 1839, 39).⁴³ Other associations were unable to prevent free-riding and declined owing to their inability to collect fees.⁴⁴

But the problems associated with prosecution associations have to be seen in context. The new police faced arguably still more serious problems. A full consideration of the foundation of the first police forces in England is beyond the scope of this article. Nevertheless, it is important to note that while there was resistance to the Metropolitan police force established by Robert Peel in 1828 was modest, ‘[m]ost of the backbenchers who acquiesced in the appearance on London’s streets of blue-uniformed and truncheoned ‘Peelers’ would have been mortified had they known that these strange novelties formed the advance guard of a professional police force which would appear throughout the country within thirty years’ (Evans 2001, 244). Many areas of the country resisted the introduction of police forces.⁴⁵ In Yorkshire in 1839, for example, magistrates were ‘bombarded with petitions from numerous townships whose inhabitants, like those from Kirby Moorside, had ‘learnt with great Alarm, that the Establishment of a Rural Police is intended’ despite the fact that in their opinion there were ‘no instances of insubordination or riotous Proceedings … but that on the contrary, the Riding is in a state of perfect tranquility that rendered a police force ‘altogether unnecessary and uncalled for” (quoted in Taylor 2002, 26).⁴⁶

Prosecution associations persisted after the introduction of professional police forces across England between 1828 and 1856. They worked in cooperation with the new police.⁴⁷ Many associations continued to pay rewards to policemen into the 1880s or later, just as they had previously paid constables. However, in the long-run the emergence of the police robbed prosecution associations of their principal reason for existing, particularly as the police began to take over the burden of prosecuting felons from the 1870s onwards; those societies that survived did so mainly on the basis of their success as convivial societies.⁴⁸