Criminal numbers
Historians have expended considerable energy on the collection and analysis of criminal numbers. For anything before 1805, they are reliant on the committals for trial on indictment at the courts of quarter sessions and assizes. Indictable offences were those for which an offender had to be tried before a professional judge or bench of magistrates, assisted by a jury, the latter handing down the verdict, the judge or magistrates pronouncing sentence. By and large, all the more serious offences â homicides, sexual offences, robberies, housebreaking and burglary, forgery, fraud, aggravated assaults, and larcenies â were tried on indictment. Excluded were the minor categories of offence, such as vagrancy, drunkenness, breaches of the licensing and game laws, and the less serious assaults and breaches of the peace. These latter offences were tried summarily by magistrates sitting without a jury at petty sessions. Records of petty sessions have not usually survived, except occasionally in the estate papers of gentlemen justices. Quarter sessions, held in the county town four times a year (or more in some counties where crime was common), tried more of the indictable offences than assizes did, especially larcenies, housebreaking, assaults, and robberies without serious violence.1 The numbers derived from these legal documents have been combined into statistical time series for a number of counties and correlated with various indices of social and economic life. Whether such time series advance our understanding of criminality, as distinct from patterns of prosecution, remains in dispute to this day.2
In 1805, Parliament ordered the national collection of information on those people committed for trial for indictable offences. Annual statistical returns of crime in England and Wales were published in parliamentary papers, the returns including the number committed for trial, discharged (on âno true billâ being found), acquitted, or convicted, and the numbers of sentences imposed for each offence. The figures, which were closely scrutinized by contemporary commentators, revealed a steady increase in indictable committals to trial for England and Wales: from 4,605 in 1805 (3,267 males, 1,338 females) to 13,710 in 1820, 18,107 in 1830, and 31,309 in 1842 (25,740 males, 5,569 females), a rate far in excess of population growth.3 While the figures of the number of persons tried for indictable offences increased to seven times their original amount between 1805 and 1842, population increased by only 80 per cent. While some contemporaries sounded the tocsin to warn of moral ruin, others recognized that much of the increase came from changes in the level of prosecution. No figures for those tried summarily were yet included in the official papers, but from 1836, returns of those sent to prison for summary offences partly filled the gap. There had been a considerable extension of the magistratesâ powers of summary jurisdiction in the 1820s, and this jurisdiction was further increased by the Juvenile Offenders Acts of 1847 and 1850, and the Criminal Justice Act of 1855, which allowed minor larcenies, while remaining indictable offences, to be tried by magistrates courts. Finally, from 1857, the enlarged series known as the Judicial Statistics carried information on the number of indictable offences known to the police (the closest one can get to the incidence of âactualâ crime), the number of people committed to trial for indictable and summary offences, and the number and personal characteristics of those imprisoned on conviction.4
It should be noted that official data was also provided on capital convictions and executions over a more extended time frame than ever before. The 1819 Commons Select Committee on Criminal Laws published a statement of the number of persons who were capitally convicted, and of those who were executed, in London and Middlesex from 1749 to 1818, together with the offences for which the executions took place.5 This statement gave a depth of history to capital punishment it had not previously enjoyed.
Most historians agree that the official returns provide no consistently accurate picture of levels of crime in the past. The âdark figureâ of unrecorded crime is known to be large. Historian Peter King estimated that no more than 1 in 20, or 5 per cent, of thefts led to a prosecution.6 Victims of crime might resort instead to rewards to recover their stolen goods. Even cases that went before an examining magistrate were typically diverted from legal channels by persuading the parties to âcompoundâ the felony or make up their quarrel.7 Nor can we assume that criminal indictments represented a constant fraction of actual crime. The justice system created no mirror image of âactualâ crime; it dipped into the reservoir of unreported crime, and it plucked out a selective number of offenders. For most of the period under examination, it was the responsibility of the victim who had been burgled or assaulted to bring a prosecution, gather the evidence, and plead his or her case before the court.8 Many victims chose not to prosecute: they were unwilling to incur the loss of time and expense that mounting a prosecution involved; they were reluctant to inflict the severe penalties, notably capital punishment, prescribed by law for crimes of property. As King points out, even small changes in victimsâ willingness to prosecute, stimulated by alarmist newspaper reporting, or repressed by the availability in wartime of diverting offenders into military service, would leave a marked impression on the statistical record.9 And it was not just prosecutors who could make a difference. When a region seemed engulfed by crime, grand juries sent more defendants to take their trial, petty juries acquitted fewer defendants (and judges sentenced more severely).
Changes in the law could also have an appreciable effect on the statistical record. The Juvenile Offenders Act, 1850, and the Criminal Justice Act, 1855, which allowed previously indictable larcenies to be tried summarily by magistrates, revealed how many potential prosecutors had been deterred from prosecuting. As David Philips showed, the effect of the Acts was not alone to transfer the larcenies which had previously been dealt with in quarter sessions to petty sessions, but also to âincrease by a vast extent the number of such larcenies which were tried at all.â10 At the national level, trials on indictment fell in the mid-1850s from 21,000 to fewer than 10,000, while summary trials rose from 3,000 to 35,000. On the other side of the ledger, however, is the fact that thousands of workingmen were clearly not deterred by the expense from prosecuting other workingmen for small thefts, suggesting that one can also exaggerate the size of the dark figure of crime.11
Handling criminal numbers
There is a general consensus among historians that there are three different ways of handling these aggregate data sets, which they categorize as âpositivist,â âinter-actionist,â and âpessimist.â12 The first approach was pioneered by Douglas Hay, among other early-modern historians, who used indictment-based time series; and by Vic Gatrell and Tom Hadden, who used official nineteenth-century statistics.13 They examined fluctuations in crime, particularly in crimes against property in the short term, over, say, 2 to 10 years; and plotted the connections between measurements of crime and movements in the price of food or business cycle â to determine if crimes against property were a response to material deprivation. They proceeded on the assumption (which others find questionable) that changes in the number of indictments are a reliable guide to changes in the far larger number of offences which went unsolved or unprosecuted. They do not argue that the numbers of committals represent directly the numbers of illegal acts, only that the movements in the numbers of committals reflected the same movements in the numbers of illegal acts. The second and more widespread approach taken to the official statistics, first employed by David Philips on court data drawn from the Black Country between 1835 and 1860 (some 20,000 criminal cases), has been to eschew judgments about crime rates and focus instead upon what the statistics of crime reveal about the changes in law and law enforcement.14 âControl wavesâ were as likely to affect the criminal statistics as âcrime waves.â15 This is to use crime figures as an index of the activity of the criminal justice system; to uncover the interaction between the number of offences, on one hand, and the willingness of the public to report and prosecute crime, or police willingness to record and act upon public information, on the other hand.
The third and final approach is the pessimistic position, taken by Rob Sindall among others, that the official figures are so distorted, they are no more than a guide to what Victorians believed was the state of crime, and to what âmoral entrepreneursâ employed in their alarmist quest for police or prison reform. The most extreme version of the pessimist position appeared in Howard Taylorâs argument that the crime statistics were knowingly fabricated by the police by the under-recording of crimes or by manipulating prosecution practice, a premeditated conspiracy of over one hundred separate police forces, under the pressure of fiscal and manpower constraints, and the desire to improve clear-up rates (or the proportion of reported crimes detected by the police).16
Thus, positivists believe that crime statistics can unravel the time trends in crime rates; interactionists believe that the statistics can uncover the workings of the criminal justice system; and pessimists insist that the statistics are too flawed to be useful â apart, that is, from defining the burden of crime that Victorians believed they confronted, and the role such public perceptions played in shaping responses to crime. The latter is a vital consideration. A number of the documents in this volume testify to the importance, in John Beattieâs phrase, of âwhat contemporaries thought changes in the levels of indicted crime actually meant.â17