Edward Banfield

The Unheavenly City

Most stealing is done by persons who want small amounts now. For them a job that must be worked at regularly and that pays only at the end of the week is not a real alternative to stealing. Even if the wage rate is high, such a job is of no interest to one who wants only a few dollars—enough, say, to buy a couple of six-packs of beer and a carton of cigarettes—but wants them now—this very day, perhaps this very hour. What is needed to reduce stealing, then is not so much high employment and rising incomes as it is greater opportunity for people who live in the present to get small sums when they want them. Paying unskilled workers by the day instead of the week would help matters some. So would paying them for days that they are prevented from working by weather, the illness of a family member, or some other good reason. So far as employment levels are concerned, the need is not so much for more 'good' jobs as it is for more casual ones—jobs that, although not high paying, are readily available to persons who want to 'make a few bucks' when, and only when, the spirit moves them. Boys, especially, need such job opportunities; perhaps there is no more economical way of reducing juvenile delinquency, and thus crime in general, than by repealing the minimum wage and relaxing the child-labor and school-attendance laws. The main effect of these laws, it is probably safe to say, is to make stealing the boy's easiest, if not his only, way of getting the money and other things he thinks he needs.

If making it easier to earn money is one way of influencing the outcome of the individual's calculus of profit and loss when he contemplates crime, increasing the probability both of his being caught and of his being severely punished is another. Obviously, one way of doing the former with respect to the kinds of crimes that occur in public places is to put more policemen on patrol. From the standpoint of the calculating individual, the greater probability of a policeman's appearing represents a cost; the higher this cost is, the less likely (other things being equal) the individual is to commit a crime. 'Operation 25,' an experiment tried all too briefly by the New York Police Department in 1954, gave indications that rather dramatic results can be obtained by 'saturating' an area with patrolmen. For a four-month period the force in a high-crime district of Manhattan was doubled. Muggings fell from 69 to 7, auto thefts from 78 to 24, and assaults from 185 to 132. Murders, however, increased from 6 to 8. (Since it generally takes place in private, murder is not likely to be deterred by the presence on the street of any number of policemen.) To what extent the drop in crime reflected not deterrence from crime altogether but rather deterrence from crime in that particular precinct (that is, its displacement to a less heavily patrolled one) there is, of course, no way of knowing.

Attaching a stiffer penalty to an offense sometimes, but by no means always, raises the cost of the action significantly. When the penalty for prostitution was drastically reduced in New York City, the number of prostitutes in the city increased very quickly, some coming from cities hundreds of miles away in order to take advantage of the lower costs of doing business there. As a rule, however, the opportunities to deter crime by threatening severe penalties are very limited. One reason for this is that judges and juries will not enforce laws that carry penalties they consider 'out of line.' Another is that if the probability of being caught is negligible, it makes practically no difference to the calculating individual what the penalty is. In fact, the probability of being caught is very often negligible. (For a Negro youth gang member, a study estimated, the probability of arrest for involvement in an instance of potential violence was about .04 if he was of average skill and about half that if he was very skillful.) According to a recent study, certainty of punishment has a greater effect than severity on most crime rates. For all offenses except homicide, the rate of crime decreases as the level of certainty increases, regardless of the level of severity. Curiously, the association between certainty of punishment and a low crime rate is weakest in more highly urbanized states.

It will be seen that the threat of even very stiff future penalties will not have a deterrent effect upon radically present-oriented individuals. It is likely that even to a normal person a punishment appears smaller the farther off in the future it lies. With the radically present-oriented, this distortion of perspective is much greater: a punishment that is far enough off to appear small to a normal person appears tiny, or is quite invisible, to a present-oriented one. His calculus of benefits and costs is defective, since benefits are in the present where he can see them while costs are in the future where he cannot. Accordingly, even if he knows that the probability of his being caught is high and that the penalty for the act is severe, he may commit the crime nevertheless; no matter how severe, a penalty that lies weeks or months away is not a part of reality for him.

The implication is that in order to deter juveniles and lower-class persons (the present-oriented classes of offenders), penalties must follow very closely upon the commission of crimes. Speeding up court processes so that fines will be imposed or jail sentences begin hours or days, rather than weeks or months, after arrest would probably reduce somewhat the rate of common crimes even if the probability of arrest and conviction remained as low as it now is.