In 1895 Emile Durkheim made the startling observation that crime is normal in a society, ‘an integral part of all healthy societies.1’ This view was novel then and it is still all-too-novel now, although recent concern about violence in America and elsewhere has brought an increasing awareness of the continuity of disorder in our own history.
The tardiness of this awakening is unfortunate. Delinquent behaviour must be considered normal to a society for reasons that by now are becoming obvious. As far as we know, no modern society exists which is crimeless. Moreover, numerous studies have demonstrated that most of us, regardless of our social standing, are involved in delinquent behaviour at some point in our lives.2 Crime, like the poor, is apparently always with us. But Durkheim went farther than this, arguing that crime should be considered normal because it plays a useful role in the integration of a society. Delinquent behaviour, he suggested, is boundary-maintaining: by injuring public morality, the offender draws the community together in indignation, thereby fostering social unity and helping the community to define and affirm a common morality and identity.3 Later commentators have suggested other possibilities; some suggest for instance, that criminal behaviour is a form of social adaptation, representing an attempt to adjust to social norms, albeit in socially illegitimate ways. However, regardless of whether or not it is functional, the point is clear: crime, like sex and disease, is not an occasional aberration but a usual occurrence in human experience and as such deserves the attention of historians as well as that of criminologists. Although our knowledge is admittedly still incomplete, historical literature abounds with information about such everyday matters as birth, marriage and ploughing methods. Occasional outbreaks of collective violence – riots, revolutions, mass protests – have received even more attention. But surprisingly little is known about past patterns of criminal behaviour. It is high time that the more mundane forms of disorder received their due. For only through historical analysis can we develop a conceptual analysis of the relationship of crime to broader social processes such as industrialisation, and through this more than a superficial assessment even of present crime trends.
The Issues
This study, then deals with the history of crime – or, to put it another way, with crime as social history – in Germany and France during the nineteenth century. It does not, however, concern itself with the particular motives and situations of individual offenders. Nor does it report the fascinating details of crimes and trials. It records, instead, the results of an analysis of gross statistical patterns of crime for rather large geographic areas over long periods of time. In the parlance of the cliometricians, this is a quantitative study on the macro-level. That such an approach is taken is not meant to imply that closer case histories are unnecessary or illegitimate forms of history; these approaches are in fact desperately needed. Rather, the present approach was adopted in the belief that an analysis of this sort was a necessary first step to an understanding of the crime phenomenon. For in spite of many attempts, both quantitative and otherwise, to explore historical patterns of crime, surprisingly little is known about the behaviour of past crime rates and their overall causes. Did crime rise or fall during the past century? Were rates lower in the nineteenth century than later? Which areas had the highest rates? Which the lowest? In what periods were crime rates especially high? Such questions are by no means new. Indeed, they were raised by A.M. Guerry, often considered the founder of modern criminal science, as early as 1833 and they have been widely discussed since then, but the results have been rather contradictory and unconvincing, due, in many cases, to imprecise methods or indexes and to uncritical attitudes towards sources.4 Until such basic questions as whether the incidence of crime increased or decreased during the century are answered, we can say little about the causes and meaning of crime.
Thus the first task is strictly empirical and requires the construction of usable and reliable indexes of crime. To this end, some new materials and methods will be used and, more importantly, attention will be given to deficiencies in crime records; if distortions in crime statistics cannot always be eliminated, they can at least be recognised and minimised. Only after this descriptive work is done can we move to issues of a more analytical nature. Three specific problems provide the focus here. None of them is new, but each merits further attention because of the tenuous and contradictory nature of many earlier conclusions. Moreover, in each case, if satisfactory answers to the questions can be suggested, important insights into nineteenth-century society – as well as into the the nature and significance of criminal behaviour – may be gained.
The relationship between crime and economic conditions serves as one basic focus. Did crime rise in periods of depression? Did crime decrease with prosperity and rising standards of living? The world-wide depression of the 1970s makes these questions especially pertinent, but the issue is old and has been debated heatedly. Because the issue is both important and controversial, therefore, it merits further investigation using more varied and precise indexes and methods than have usually been employed.
The second and third problems get at some of Western man’s most basic assumptions and prejudices about both crime and modernity. Everyone knows that crime is more frequent today than it was in the stable rural milieu of our grandparents and great grandparents. In fact, many would argue, such a trend is inevitable. Modernity implies a decline in respect for conventions, a reduction in social controls, a lessening of appreciation for the rights and property of others. What could be more logical than that delinquency should accompany the modernisation process? Moreover, the growth of cities is usually considered a major catalyst in this development, for during the past century popular opinion, nourished by studies which were frequently biased against the city in favour of a nostalgic view of a lost rural paradise, has associated urban life with high crime rates. The city, in the popular view, is characterised by instability, impersonal relationships, social disorganisation and weakened social controls; it is the paradigm of modern society. Consequently the city is also characterised by high rates of delinquency and the urbanisation process, which dissolves the more stable traditions of the countryside, results in crime.
These assumptions are deeply ingrained in us, even when the most liberal contemplate a trip to New York or London’s East End, but recent theory and research in various parts of the world is undermining many of them. Thus our second two questions are especially relevant here. First, were there in fact urban and rural differentials in nineteenth-century criminal behaviour; that is, can specifically rural and urban patterns of crime be isolated? Is crime really more common in the city than in the countryside and is it different in type? The second question is closely related to the first. What, if any, relationship existed between patterns of delinquency and the social upheaval which accompanied industrialisation and urbanisation? In the current terminology of the social sciences, what connection may be seen between criminality and the modernisation process? This latter question has in turn two dimensions. The most obvious, of course, is whether the social upheavals of the period led to increased crime as many assume. Just as important, however, is the question of whether patterns of criminal activity changed along with the modernisation of society. Did new types of criminal behaviour emerge, paralleling the modernisation of social relationships and values? The question of modernisation, then, is not simply whether social change led to increased crime via a breakdown of social controls and traditions (or by some other mechanism), but also whether more modern values and patterns of behaviour are reflected in delinquent behaviour, that is, whether the nature of crime and its probable motives changed regardless of rate.
Nineteenth-century Germany and France provide an excellent context for the examination of these questions.5 Between 1830 and 1914 both countries experienced substantial urbanisation and industrialisation. The rise of large cities and of large-scale, mechanised industry represented a significant break with the past and, in many cases, a massive disruption of traditional structures, values and routines. Centuries-old methods of production were abandoned and replaced by newer, more modern methods. Machines often took the place of manual labour, thereby displacing many persons and altering required work skills as well as the pace of work itself; although domestic industry survived well into the nineteenth century and even was temporarily boosted by rising demand in some instances, the characteristic work arena shifted from the family and small artisanal shop to the large factory where numerous workers, aided by machines, laboured under strict supervision. Populations grew rapidly. Increasing numbers of the rural population, pushed by the lack of jobs and land in the countryside or pulled by the lure of opportunities in the city, moved into new urban surroundings to seek work in factories and mills which were located there. Consequently, during the nineteenth century the percentage of the population which resided in urban areas in these two countries increased greatly; cities of unprecedented size arose; and the industrial city was born. Although standards of living generally rose in the long run, they fell for some elements of the population – especially at first – and were subject to frequent fluctuations: the nineteenth century was punctuated with economic crises of varying intensity which brought danger of actual starvation to some of the population and meant, at the very least, uncertainty for others. Hundreds of thousands of people were affected by these phenomena, many adversely, and even where actual change or hardship was not experienced, the threat of change itself caused fear and discontent. Both Germany and France, in other words, proved an important testing ground for the impact of the industrial revolution on crime.
Despite the difficulties involved, and in contrast to most historical studies undertaken, it was essential to adopt a comparative approach to answer the three questions we are addressing. Statistics from a single place and period are often quite uninformative and even misleading; it is only through comparisons among patterns in a variety of situations that significant and convincing conclusions are possible. We have seen claims, for example, that crime declined in England as industry matured. This is in fact debatable, but the larger point is that, dealing only with one, possibly special case, such claims tell us nothing about general relationships between crime and modern social change.6 Here is where Germany and France provide important correctives from the comparative standpoint. The two nations shared common experiences during industrialisation and urbanisation in the nineteenth century: both nations underwent substantial social and economic changes, the populations of both were affected by severe economic fluctuations, and a significant proportion of both populations lived in cities by the. latter part of the century. Moreover, both participated in the same Western cultural system and therefore cultural differences, although significant enough to be interesting, are not overwhelming. Yet, upon closer examination, the differences between the two countries are many. France began the process of industrialisation and urbanisation earlier than did Germany, yet proceeded much more slowly and was less completely transformed by 1914.7 Conversely, the process began later but was telescoped into a shorter period in Germany. Heavy industry was more important in German industrialisation than in French, a factor which directly affected the extent of urban and industrial concentration and the nature of the working class. In spite of rapid and massive urbanisation, however, rural residence for some industrial workers persisted longer in Germany than in France due, in part, to a well-developed railway system. The role of the Government in stimulating industry, and also in controlling as well as protecting workers, was more pronounced in Germany than in France. Likewise, there were significant differences in working conditions, labour agitation, and in the size and efficiency of repressive apparatuses in the two countries.
Such differences, in spite of basic similarities, make these two national units particularly useful for comparative purposes. But for several reasons great care must be exercised when comparing national units. For one thing, differences in legal, judicial, and repressive systems complicate comparisons. Thus, while comparisons of trends in crime in the two countries are possible, variables such as how crimes were defined, what evidence was required for reporting or arrest, or how crimes were actually recorded make some international comparisons of crime levels tentative at best. Secondly, not all areas within either country shared similar experiences. Some regions in both Germany and France, such as the Ruhr area in Germany and the Nord and Pas-de-Calais in France, industrialised massively and rapidly during the century while other areas such as East Prussia or the Finistère were virtually untouched by these processes. Thus simple national comparisons may obscure important regional differences.
This raises one further issue which must be considered, if only secondarily. The regional diversity of these countries, noted above, suggests that the value of national rates as indicators of criminal patterns needs to be tested. Even though most questions are framed, and answered, on a national level in this analysis, we also test the extent to which national rates actually provide useful information about the incidence of delinquency. Of course, national rates may actually reflect behaviour which is distinctly national in character, attributable to culture or other shared experiences; German patterns of delinquency may be different from French. On the other hand, though, such rates may be purely mathematical averages of quite disparate regional or local patterns; this would be true, to return to a vital question raised earlier, if urban or rural patterns transcend national boundaries. National rates thus only have meaning if regional rates tend to be grouped around the national mean; if local rates are highly dispersed, national averages are at best mathematical abstractions. Fortunately, Germany and France in the nineteenth century also provide ample smaller regions where similarities and contrasts may be tested on a comparative basis.
The Data
A failure to recognise the significance of delinquency has not been the only cause for the neglect of crime in historical literature. Social historians have also avoided the subject of crime due to a deep distrust of crime statistics. Such suspicion is not entirely unwarranted; indeed, anyone acquainted with the controversies generated by the FBI’s soaring index of crime in America cannot help but question the validity of crime records. Deficiences in FBI figures are only partially due to inadequacies inherent in the records themselves: crime statistics, like most statistical indexes, offer innumerable possibilities for misinterpretation and misuse. Nevertheless, obvious shortcomings in contemporary crime records have caused many social scientists to regard any attempt to study crime rates in the past as futile.8
The biases and distortions in crime records must not be glossed over. However, a close study of various crime indexes from both the nineteenth and twentieth centuries indicates that nineteenth-century criminal statistics are more plentiful and provide better indexes of delinquent behaviour than is commonly assumed.9 Crime indexes are not random variables; records from the nineteenth century yield recurrent and comprehensible patterns of crime which cannot be attributed simply to biases in the records or to the activities of the agencies which compiled them. However, indexes must be carefully selected and constructed, and a recognition of basic distortions is essential.
The pros and cons of historical crime records need not be surveyed in detail here, but several guidelines for the use of crime records can be noted. Modern scholars are fairly well agreed that police statistics – and especially records of crimes made known to police as opposed to more arbitrary statistics of persons arrested – provide the best indexes of delinquency because they eliminate fewer crimes (the so-called ‘dark figure’) than any other category of record. Although it is often assumed that such statistics are lacking for the nineteenth century, a search of archival materials has turned up usable police statistics in some surprising places (e.g. in city administrative reports). Nevertheless, in practice a compromise is often necessary; most European governments opted for court records during the nineteenth century and thus a large body of well-organised and carefully collected court data is available on a national basis while police indexes are diverse and localised. Fortunately, comparisons of court and police data suggest that for some purposes court records may not be as bad a source as might be expected, so long as certain qualifications are recognised. Most importantly, the dark figure in nineteenth-century court records was not only larger than for police records, but also for certain property crimes such as theft it increased during the course of the century due, at least in part, to the failure of law enforcement agencies to expand as rapidly as population and crime levels and also, perhaps, to a decreasing emphasis upon the sanctity of private property which may have accompanied the urbanisation process.10 Thus not only do court records significantly underestimate actual levels of crimes, but trends in theft rates based upon court records are much too negative, causing increases in theft rates to be underestimated or even ignored altogether.11
A second caution is that actual crime categories to be examined must be selected discriminately..Many crimes, by their nature, result in poor indexes. Sex crimes (where the victim ma...