An understanding of the interaction of the social and physical factors that create high crime rates in low- and moderate-income housing developments is useful not only for devising remedies to solve their problems, but also for developing strategies for stabilizing neighboring communities composed of single-family housing.
Fig. I-17 shows the influence of different social and physical factors on the crime rates in low- and moderate-income projects operated by the New York City Housing Authority. This analytical technique, called “stepwise regression analysis,” is employed when many different factors interact to produce a particular effect (e.g., a rise in crime rates). The technique isolates those factors which contribute to the effect most strongly and independently of other factors. In Fig. I-17 the percentage of population receiving welfare is shown to be the most important factor, followed by building height (or the number of families sharing the entry to a building).
Fig. I-17: Crime rates as explained by social and physical variables.
Those social variables that correlated highly with different types of crime also correlated highly with each other. These include: the percentage of resident population receiving welfare (excluding the elderly); the percentage of one-parent families receiving AFDC; and the per capita disposable income of the project’s residents.
My interviews with residents, management, and police provide the following explanation for the correlation of these social factors and crime rates: that a one-parent household headed by a female is more vulnerable to criminal attack; that families with only one adult present are less able to control their teen-age children; that young teen-age AFDC mothers are often victimized by their boyfriends; that the criminal activity of the poor is tolerated, if not condoned, among the poor; that the poor, and particularly the poor of racial minorities, are unable to demand much in the way of police protection; and that the commission of crime against residents in ghetto areas requires minimal skill and risk.
The physical factors that correlate most strongly with crime rates are, in order of importance: the height of the buildings which, in turn, correlates highly with the number of apartments sharing the entry to a building; the size of the housing project “or the total number of dwelling units in the project”; and the number of other publicly assisted housing projects in the area.
The above suggests that there are two classes of physical factors that contribute to crime rates: 1) those such as “project size” or the “number of publicly assisted projects in the area” which reinforce social weakness and pathology; and 2) those such as “building height” or “the number of units per entry” which affect the ability of residents to control their environment. The first class of physical factors may also be considered another class of social variable: For instance, if certain social characteristics such as the percentage of AFDC families correlate highly with crime rate, then we can anticipate that a large number of such families gathered together in one area may aggravate the crime problems still further and increase the crime rate per capita.
The significance of this is not simply that the presence of more potential criminals creates proportionally more crime, but also that a concentration of potential criminals actually increases the rate of crime. Thus, large low-income projects, or low-income projects surrounded by other low-income projects, suffer a higher crime rate than small or isolated projects even when the percentage of AFDC families remains the same in all the projects.
A frequent complaint of residents of communities surrounding large public housing projects is that the teen-age criminals living in the projects make use of the large, anonymous environment of the housing project as a place to run back to and hide in. For example, there is a particularly notorious project in Jersey City, which is located adjacent to U.S. Highway 1 entering New York City. A traffic light at an intersection that borders the project forces truckers to stop there on their way into New York. Teen-age project residents have developed a pattern of hijacking trucks at the stoplight, throwing the driver out, and then driving the truck into the project. The truck is then emptied in a matter of minutes and the loot hidden in vacant apartments.
The relationship between the socioeconomic characteristics of residents and a project’s crime rate had long been suspected. The most fascinating new information to come out of our 1970 analysis, therefore, was that of the influence of building height and number of units per entry in predicting crime rate. Regardless of the social characteristics of inhabitants, the physical form of housing was shown to play an important role in reducing crime and in assisting residents in controlling behavior in their housing environments.
In addition to the fact that buildings with a larger number of families sharing an entry experience higher crime rates than those with few families per entry, they are also vulnerable to additional types of criminal activity. Most of the crime experienced by residents of single-family buildings is burglary, committed when members of the family are either away from home or asleep. By contrast, the residents of large, multifamily dwellings experience both burglaries and robberies (muggings). The higher crime rate experienced by residents in large multifamily dwellings is mostly attributable to the occurrence of robberies in the interior common-circulation areas of multifamily buildings: lobbies, hallways, stairs, and elevators. These are also the areas where criminals wait to approach their victims and force them into apartments for the purpose of robbing them.
Of a total of 8,611 felonies reported in all New York City Housing Authority projects in 1969 (excluding intra-household incidents), 3,786, or 44 percent, were committed in the interior public areas of buildings. Of the crimes committed in interior public areas, 3,165, or 84 percent, were robberies. The breakdown by location of the felonies taking place in interior public areas was: elevators, 41 percent; hallways, 22 percent; lobbies, 18 percent; stairways, 9 percent; roof landings, 2 percent; and other, 8 percent.
Although the socioeconomic characteristics of the residents exert a strong influence on crime rate, the physical characteristics of the buildings and the project can exert a counteracting influence. The physical form of residential environment can, in fact, ameliorate the effect of many of the problems created by the concentration of low-income one-parent families with teen-age children.
The more complex and anonymous the housing environment, the more difficult it is for a code of behavior following societal norms to become established among residents. It is even difficult for moderate-income families with two adult heads of households to cope with crime and vandalism problems in poorly designed environments, but when poor and broken families are grouped together in such a setting, the results are nothing short of disastrous. The public housing projects now experiencing the highest vacancy rates are those which consist of the worst mixture of social and physical attributes.
Fig. I-18: Variations in crime rate as produced by different socio-economic groups occupying different building types.
Fig. I-18 compares the vulnerability to crime of low-income one-parent families in different building types with the experience of moderate-income two-parent families living in the same building types. These are the further results of our 1970 analysis of New York City Housing authority data. It shows that low-income one-parent families are more vulnerable to poor building design than moderate-income two-parent families. Although two-parent moderate-income families suffer higher crime rates in high-rise buildings than they do in walk-ups, the crime rate does not increase as dramatically with building height as it does for low-income families. Moderate-income two-parent families living in twelve- to thirty-story buildings experience a lower crime rate than low-income one-parent families living in six- and seven-story buildings.