Introduction : Crime Mapping and Crime Prevention

Crime maps have only recently begun to emerge as a significant tool in crime and justice. Until a decade ago, few criminal justice agencies had any capability for creating crime maps, and few investigators had the resources or patience to examine the spatial distribution of crime. Today, however, crime mapping is experiencing what might be termed an explosion of interest among both scholars and practitioners. This introduction begins by examining some early examples of mapping of crime, focusing in particular on factors that inhibited the widespread integration of mapping into crime prevention research and practice in the past. It then turns to innovations in mapping technologies and crime prevention theory that have recently brought crime mapping to the center of trends in crime prevention. The final section introduces the contributions that follow and discusses how they illustrate the many uses of mapping in crime prevention. It examines the pitfalls and problems that researchers and practitioners are likely to encounter in developing and analyzing maps, and the potential advances in crime mapping we might expect in coming decades. Address correspondence to: David Weisburd, Institute of Criminology, Faculty of Law, The Hebrew University, Mt. Scopus, Jerusalem, Israel, or Tom McEwen, Institute for Law and Justice, Inc., 1018 Duke St., Alexandria, VA 22314. 2 — David Weisburd and Tom McEwen Figure 1: Map of Cholera Deaths and Locations of Water Pumps Crime Mapping and Crime Prevention — 3 In London in the nineteenth century, cholera was a fearful disease that raged in epidemic proportions and left death and suffering in its wake. The disease seemed to be concentrated in specific neighborhoods. In one epidemic, a physician decided to identify with precision where in the city the deadly disease left its mark, with the hope of finding some pattern to its destruction. Dr. Snow plotted the location of deaths from cholera on a map of central London in September 1854 (see Figure 1). Drawing from a theory that contaminated water causes cholera, he also marked the locations of the area's 11 water pumps. He analyzed the scatter of dots on the map and noticed that they concentrated near the pump on Broad Street. While Dr. Snow suspected the water pump was the problem, he was not entirely certain because one rectangular area near the pump showed no deaths and an area near another pump indicated a second, smaller concentration of deaths. After further inspection, he found that the rectangular space was a brewery where employees did not drink water because they were provided with free beer. Investigation of the other area determined that residents had friends and relatives near the Broad Street pump, and when visiting they often took jugs with them to fill up because the water there seemed to taste better than that from their local pump. Dr. Snow had the handle of the Broad Street pump removed, and the cholera epidemic came to an abrupt halt after having taken more than 500 lives. Dr. Snow's efforts provide a dramatic example of the use of mapping for informing public policy. But maps themselves have a long history as a basic form of human communication. They have been used to navigate streets and oceans, to portray trends in weather or population, to illustrate political divisions or military strategies, to define boundaries or to reinforce them. For example, a Tahitian native communicated his knowledge of South Pacific geography to Captain Cook by drawing a map, thereby illustrating that the islanders were quite familiar with the idea of mapping. Real estate maps dating to 2000 B.C. found in Mesopotamia and Egypt illustrate the fact that maps are as old as human civilization. In fiction, detectives often look to maps to untangle complex clues or to bring together seemingly disparate events. A serial murderer may be caught in part because of the clustering of kidnappings in particular types of places. The mystery of the whereabouts of booty from a bank robbery, hidden for half a century, may be unraveled by plotting the locations of cemeteries that lie close enough for a stash to be made before the criminals were caught (Grafton, 1995). Based on what we see in movies and on television, we might expect to see the 4 — David Weisburd and Tom McEwen operations room of a police department laced with colored pins maps. But in practice, crime maps have only recently begun to emerge as a significant tool in crime and justice. Until a decade ago, few criminal justice agencies had any capability for creating crime maps, and few investigators had the resources or patience to examine the spatial distribution of crime. Today, however, crime mapping is experiencing what might be termed an explosion of interest among both scholars and practitioners (see, e.g., Block and Dabdoub, 1993; Eck and Weisburd, 1995; Harries, 1990). Crime mapping has suddenly emerged as a major tool in crime prevention. In developing this volume, we sought to bring together scholars, crime analysts, and practitioners on the cutting edge of both research and practice in mapping crime. At a time when new mapping technologies are just beginning to be integrated into crime prevention, we wanted to provide examples of how maps could be used in developing policy and theory, and to illustrate the prospects and problems that crime mapping presents. Our volume includes contributions that examine mapping in real life criminal justice contexts, as well as examples of new technologies and future trends that have yet to be implemented in practice. We are concerned with crime prevention theory and crime mapping technologies. Our choice of such a wide range of topics is not accidental: successful mapping of crime demands an integration of theory and data, as well as a practical understanding of the real life context of crime and justice. In introducing our volume, we think it important to provide historical perspective to the development of crime mapping in crime prevention. We begin by examining some early examples of mapping of crime, focusing in particular on factors that have inhibited the widespread integration of mapping into crime prevention research and practice. We then turn to innovations in mapping technologies and crime prevention theory that have recently brought crime mapping to the center of trends in crime prevention. Finally, we discuss how the contributions that follow illustrate the many uses of mapping in crime prevention, the pitfalls and problems that researchers and practitioners are likely to encounter in developing and analyzing maps, and the potential advances in crime mapping that might be expected in the coming decades. CRIME MAPPING: EARLY APPLICATIONS The idea of mapping crime is not new and, in fact, dates back to the early 1800s in France. A review of the historical literature from Crime Mapping and Crime Prevention — 5 that period to the present time shows several epochs during which interest in crime mapping was great, but then faded dramatically. In this section, we examine three such periods and discuss the reasons why what seemed like promising beginnings did not lead to sustained interest in crime mapping. In 1829, Adriano Balbi and Andre-Michel Guerry created the first maps of crime (Kenwitz, 1987; Beirne, 1993). The collaboration itself is of interest because it combined Balbi's training in ethnography and general mapping techniques with Guerry's training as a lawyer interested in patterns of criminality. Using criminal statistics for the years 1825 to 1827 and demographic data from France's latest census, they developed maps of crimes against property, crimes against persons, and levels of education. Comparing these maps, they found that the northeastern portion of France (from Orleans to the FrancheComte) was better educated, that areas with high levels of crimes against property had low incidences of attacks on people, and that the areas with more property crime were populated by people with higher levels of education (see Figure 2). While the results regarding geographic differences in educational levels came as no great surprise, the others ran counter to popular views at the time. Guerry, however, paid little attention to these reactions because he was not interested in developing or testing theories (Oberschall, 1989). The Belgian astronomer and statistician Lambert-Adolphe Quetelet attempted to fill the theoretical void. In 1831 and 1832, he independently published three maps dealing with the same themes but spreading across larger areas. Quetelet saw a correlation between crime and several variables including transportation routes, education levels, and ethnic and cultural variations. Quetelet continued theoretical development through his concept of the "average man" and his quest to discover, through statistics and "social physics," the explanation of societal behavior (Quetelet, 1835). His contributions to statistics, which were very controversial at the time, suggested the application from astronomy of the normal distribution and error measurement to social phenomena (Maltz, 1991). Quetelet's use of statistical tools combined with the average man concept was founded on the belief that aggregations of data provide statistical stability, assuming there is no change in any underlying causal relationships. As he stated, "The greater the number of individuals observed, the more do individual peculiarities, whether physical or moral, become 6 — David Weisburd and Tom McEwen Figure 2: Balbia and Guerry (1829) Maps Comparing Crime and Instruction effaced, and allow the general facts to predominate, by which society exists and is preserved" (Quetelet, 1835:12, as reported in Stigler, 1986). Indeed, Quetelet found stability over time in crime, birth, and suicide rates and other social phenomena, to the extent that critics said he was questioning the very existence of free will. Despite groundbreaking work in providing explanations for the distribution of crime, these ecological perspectives were hastily discarded with the advent of a "positivist" criminology eager to locate the Crime Mapping and Crime Prevention — 7 causes of crime within the biological and physiological framework of individuals (Morris, 1958; Beirne, 1993). Robinson (1982) draws the following conclusion about the development of thematic maps and Quetelet's (1835) statistical approach: Although thematic maps of moral statistics continued to be made [into the 1860s], especially of instruction, their developmental period had run its course. Subsequent attention seems to have been oriented more toward sociological interpretation and away from geographical variation. In a sense this reflects a greater concern with Quetelet's provocative ideas of "social physics" and a lessening of interest in the investigations of regional differences, such as by Dupin, Guerry, and Fletcher [Robinson, 1982:170]. There was also a very practical reason for moving away from the use of maps and regional variations in social and moral statistics. It required a considerable amount of time and effort to collect data, summarize by appropriate areas, and manually create the maps. Whatever the reasons for the demise of interest in mapping of crime, the lesson in our context is that the efforts during this period were based on reasonably good data (as collected by France's censuses), but were weak in other areas. Maps of interest could not be developed on any regular basis; crime theories were not adequately developed; and techniques for analysis were slow, time-consuming, and cumbersome. The history of mapping in the U.S. stands in sharp contrast to what has just been discussed. We sometimes forget that America is a comparatively young nation. In the early 1800s, many parts of the U.S. were years away from collecting crime and demographic data on a routine basis. A search of the police literature uncovered no references even in the first part of this century to the 19 century mapping efforts in England and France. Instead, what begins to emerge are occasional references to "spot maps" in which pins are physically placed on large street maps. These spot maps were used first with traffic accident data, which predated systematic collection of crime data. More sophisticated maps were developed by a group of scholars associated with the University of Chicago in the 1920s and 1930s. These urban sociologists, led by Robert Park, looked to characteristics of the urban environment to explain the crime problem in American cities. They mapped crime and other social characteristics in neighborhoods in the city of Chicago. Using these maps, they illus8 — David Weisburd and Tom McEwen trated the theoretical position that crime was strongly linked to social disorganization and poverty in urban settings. For example, Frederic Thrasher (1927) superimposed the "location and distribution" of gangs in Chicago on a map of urban areas in the city (see Figure 3). He found that gangs were concentrated in areas of the city where social control was weak and social disorganization pervasive. Similar conclusions were reached by Shaw and Myers (1929) in a study of juvenile delinquency conducted for the Illinois Crime Survey. In a map that looks as if it had been generated through modern computer applications (Figure 4) rather than produced by hand, they show that the home addresses of over 9,000 delinquents are clustered in areas marked by "physical deterioration, poverty and social disorganization" (1929:652). Interest in the ecological correlates of crime faded among American sociologists in the 1930s. The confident assertion by Shaw (1929) that the "study of such a problem as juvenile delinquency necessarily begins with a study of its geographical location" (p. 5) was not heeded by those who followed him. Once the relationship between social organization and crime in urban neighborhoods had been illustrated, a new generation of researchers shifted concern to elements of social disorganization and their impacts upon individual predisposition to criminality. The tedious and difficult process of mapping crime in the pre-computer age did not appear to offer potential for new and important insights. A new generation of sociologists concerned with crime sought to understand why certain individuals, both within these socially disorganized areas or outside them, chose to commit crimes while others did not (Merton, 1938; Sutherland, 1939). This question did not demand examination of the location of crime events, but rather led scholars to focus on the motivations of offenders. The idea of automated crime mapping emerged in the late 1960s. Early applications (Pauly et al., 1967; Carnaghi and McEwen, 1970) showed the potential for visual representations of crime patterns through computer-generated maps. For example, Figure 5 uses such a map to show the distribution of larcenies from automobiles in 1967 in the ninth district in St. Louis, MO. Maps were seen as offering the potential for focusing police resources in more efficient and more effective ways. Similarly, publications on crime analysis strongly advocated automated analysis of crime (Chang et al., 1979; Buck et al., 1973) and illustrated the use of geographic analysis (Brantingham and Brantingham, 1981; Harries, 1974; Pyle, 1974). Crime Mapping and Crime Prevention — 9 Figure 3: The Place of Chicago's Gangland in the Urban Ecology Note: The shaded portion indicates the approximate location of the central empire of gangland. 10 — David Weisburd and Tom McEwen Figure 4: Home Addresses of Alleged Male Juvenile