THE STATISTICS OF CRIME
By Eugene Smith
President Prison Association of New York
[Mr. Smith read a very carefully prepared paper on the above subject at the Omaha meeting of the American Prison Association. The Review would gladly print the address in full but space admits only of certain abstracts, which follow.—Editor]
In the deplorable and chaotic condition of the very sources from which all statistical matter must be drawn, it is hopeless to look for any improvement in our census statistics, unless a radical change can be effected in state administration. The records of the police, the courts, the prisons, can be made of statistical value only by the action of the state itself; and there is apparent but one method by which the state can act to this end.
There should be established in each state a permanent board or bureau of criminal statistics, whether as an independent body or as a department of the office of the attorney general or of the secretary of state. This bureau should be charged with the duty of prescribing the forms in which the records of all criminal courts, police boards and prisons shall be kept and specifying the items regarding which entries shall be made. The law creating the bureau should direct that the forms prescribed by it should be uniform as to all institutions of the same class to which they respectively apply and be binding upon all institutions within the state.
The bureau should issue general instructions governing the collection and verification of the facts to be stated in the record; it should also be its duty, and it should be vested with power, to inspect and supervise the records and to enforce compliance with its requirements. Such a bureau might secure a collection of reliable statistical matter, uniform in quality throughout the state. Indiana is now, it is believed, the only state in the Union where such a bureau exists.
But even this result is not enough. Supposing all the criminal records within each separate state to be made uniform without the state, still they would not be available for comparison or for the purposes of a national census, unless all the states could be brought to adopt the same form and method, so that all criminal records throughout the Union could be kept upon one uniform plan. Here we encounter a serious obstacle. The diversity and conflict of state laws are crying evils of our time, universally recognized and denounced, and yet the most strenuous efforts to bring about harmonious action between the legislatures of separate states have always failed. No single statute, however skilfully drawn, proposed for universal acceptance has ever yet been adopted by all the statesof the Union. Still the statesmustact in unison upon this matter of uniform criminal records or else our statistics of crime must continue to be a national failure and a national reproach.
Not the slightest reflection can be cast upon the federal census bureau; on the contrary, when consideration is taken of the fragmentary and chaotic state records with which the census bureau had to deal, the systematic and orderly results and the general deductions embraced in the census report of 1904 must be regarded as a signal scientific triumph.
Uniformity in criminal records throughout the Union we have seen to be an imperative need. Is it a visionary ideal, impossible of attainment? If there is any means through which the ideal can be realized, it is through the agency of state bureaus of criminal statistics, such as have just been suggested. Each of these state bureaus, in preparing uniform plans and forms for its own state, would naturally place itself in touch with the national census bureau; while the national bureau would not be legally vested with the slightest power to dictate to the state bureau or to direct its action,practicallyits wide experience and grasp of the entire situation would enable the federal bureau to wield commanding influence in shaping the action of every state bureau. If the creation of efficient state bureaus, of the kind indicated, in the several states could only be secured, it is not chimerical to believe that through the dominating influence of the federal census bureau, tactfully exerted, a uniform system of statistical records relating to crime could ultimately be established throughout the United States. It is the first step that counts. If a few of the leading states in the Union could be induced to establish such a bureau; if to Indiana could be added New York, Illinois, Nebraska, and in the South Virginia, the force of example would be potent in the sister states. * * *
One exceedingly common and popular error needs special mention; a marked increase in the number of convictions for crime indicates to the public mind an increase necessarily in the volume of crime committed. In fact, it may be owing to increased activity and efficiency on the part of the police and detective officers, to greater severity and thoroughness in the administration of the courts, to a change in the economic conditions of the community, to diminished care and skill on the part of offenders in escaping detection; indeed, there are many possible factors that may have combined to produce an unusual statistical result. A slight change in the laws or methods of procedure, may cause startling statistical fluctuations.
For example, in the year 1890, the number of convictions for drunkenness in Massachusetts was 25,582; two years later, the number had fallen to 8,634. An amazing diminution of drunkenness in Massachusetts—nearly 70%? Not at all; it was owing to a new statute passed in 1891, the effect of which was that only those arrested for the third time within a year were subject to conviction.
The congestion of population in cities and the progress of invention necessitates every year the enactment of numerous statutes and municipal ordinances making certain acts, that are harmful to the public, misdemeanors (that is, legally crimes); but these acts, committed in large part through ignorance or negligence, are not essentially of a criminal nature. Statistically, they swell the number of crimes committed, but most of them are not crimes in the meaning popularly attached to that word. These considerations suggest that all attempts to draw conclusions from, and to explain the significance of the rise or fall of the statistical barometer must be conducted with extreme caution.
An error into which speakers and writers upon crime are prone to fall is that of regarding the statistics of crime as a measure of the total volume of crime committed in the country, affording an answer to the vital question: Is crime increasing? There are two fundamental facts relating to crime that must never be forgotten. First, that criminal statistics are, and must necessarily always be, confined to those crimes that are known and are officially acted upon by the police or the courts. Secondly, that there is a large number of crimes that are committed secretly and are never divulged, the perpetrators of which arenever detected, and crimes that never result in the apprehension of the offender.
The crimes of this second class cannot possibly enter into any criminal statistics and yet they form a very large part of the total volume of crime committed. It does not seem to be commonly appreciated that these unpublished, unpunished crimes, which can never be included in any criminal statistics, probably far exceed in number those that are followed by conviction and punishment. * * *
In addition to unpublished crimes, there are numerous cases where crime is committed and reported to the police, but proceed no further. In these instances, the offender may be known, but has escaped or the offender is unknown and eludes detection; in either case there is no conviction and the crime remains unpunished. * * *
Perhaps the highest value of criminal statistics consists in the light they may throw upon the practical effects produced by penal legislation, by judicial procedure and by the administration of police and detective officers. For example, within the past decade, radical changes in the administration of justice have been established in this country by laws relating to juvenile offenders, and by the extended use of the suspended sentence and probation. A question has arisen in many minds whether the severity of the penal law has not thus been unduly relaxed. It is a matter of supreme importance to know whether and how far, the tenderness of the modern law toward children serves to rescue them from a life of crime—to know whether the clemency of the law toward adults by suspension of sentence and probation promotes their rehabilitation, and to know to what class of offenders this clemency may properly be extended—to know whether these milder methods of treatment are affording adequate protection to the public or whether sterner measures of restraint and discipline may be made more effective in repressing crime.
These vital questions can receive final answer only by following the subsequent career of the offenders to whom these methods are applied and thus gaining data for statistical tabulation. In the same way, the virtue of the indeterminate sentence ought to be substantiated by the statistical test. Statistics can be made to show what class of crimes comes most frequently before the courts in a given community, and whether an increase in the severity of punishment tends to increase or diminish the number of convictions.
A movement is now in progress which may greatly widen the scope of criminal statistics. It has long been realized that many persons sentenced for crime are feeble-minded and seriously defective; mentally and physically but, within the past few years, the conviction has been growing that our penal system is radically imperfect in that it provides no adequate means for deciding whether or not a person on trial for crime is really responsible criminally. * * *