Routine activities theory has guided research designed to understand a range of phenomena, including crime trends over time, distributions of crime across space, and individual differences in victimization. In addition, researchers have considered how opportunities for crime might exist at multiple levels. For example, the characteristics of one’s neighborhood and the features of the home might influence the likelihood of burglary victimization. Researchers have used various research methods to meet these different needs. The selection of research reviewed in the following paragraphs illustrates the different methods researchers have used to test hypotheses developed from routine activities theory.
A. Using Routine Activity to Predict Crime Trends
Routine activities theory was first used to understand changes in crime trends over time. To do this, researchers examine how crime rates fluctuate over time with changes in macrolevel routine activity trends to determine whether changes in routine activities are associated with changes in crime trends. If they are, this indicates support for the theory. In their initial presentation of the theory, Cohen and Felson (1979) pointed to a shift in the structural routine activities of society to explain why urban crime rates increased during the 1960s, when the factors thought to cause violent crime, such as economic conditions, had generally improved during this time period. They argued that the dispersion in activities away from the family and household caused an increase in target suitability and a decrease in guardianship. In other words, people were leaving their households unoccupied and unguarded more frequently, as well as exposing themselves as targets to potential motivated offenders. To test this hypothesis, Cohen and Felson developed a household activity ratio to measure the extent to which households were left unattended. 1 They predicted that changes in the dispersion of activities away from the family and household explained crime rates over time, arguing that non-household activities increase the probability that motivated offenders will converge in time and space in the absence of capable guardians. Using a time series analysis, they found that the household activity ratio was significantly related to burglary, forcible rape, aggravated assault, robbery, and homicide rates from 1947 to 1974 (Cohen & Felson, 1979). Consistent with Cohen and Felson’s initial study, subsequent macrolevel studies have demonstrated that variations in society’s structural organization of routine activities are related to variations in crime trends over time (Felson & Cohen, 1980). In other words, research has generally shown that routine activities that take people away from their home tend to be associated with increases in crime rates.
B. Using Routine Activities to Predict the Distribution of Crime Across Space
Routine activities theory has also been used to explain distributions of crime across space. Unlike the research just reviewed, which examined how crime rates changed in the same place over time (i.e., the United States from year to year), this type of research examines how crime rates differ across various places at the same time (i.e., different cities in the United States during a given year). Researchers have used routine activities theory to develop testable hypotheses about why some areas have higher crime rates than others. To do this, they examine whether the routine activities of people living in places with higher levels of crime differ from the routine activities of people living in places with lower levels of crime. For example, Messner and Blau (1987) hypothesized that routine leisure activities that take place in the household will result in lower crime rates, whereas those that take people away from their households will result in higher crime rates. To test these hypotheses, they used data from the 124 largest Standard Metropolitan Statistical Areas in the United States during the time period around 1980. Specifically, they hypothesized that higher levels of aggregate television viewing would be associated with lower city crime rates, because routine leisure activities that take place in the household provide potential targets with a greater level of guardianship. Conversely, they hypothesized that a greater supply of sports and entertainment establishments will be associated with higher city crime rates, because leisure activities that remove people from their homes leave suitable targets unguarded. In general, their analyses support these hypotheses. Higher levels of television viewing were associated with lower rates of forcible rape, robbery, aggravated assault, burglary, larceny, and auto theft. Conversely, a greater supply of sports and entertainment establishments was associated with higher rates of murder and non-negligent manslaughter, forcible rape, robbery, aggravated assault, burglary, and larceny.
C. Using Routine Activities to Predict Differences in Victimization
Routine activities theory has also been used to explain differences in victimization across individuals. Although Cohen and Felson (1979) initially used the theory to explain national-level crime trends, the mechanisms described by the theory are actually microlevel in nature: A victim comes into contact with an offender in the absence of any capable controllers. This has led many researchers to use individual-level victimization data to understand differences in victimization risk given the routine activities of the potential victim. Specifically, researchers compare the routine activities of victims to those of non-victims to understand the effect of lifestyle and routine activities on the likelihood of victimization. Victimization survey data have become increasingly available in recent decades, making such methodology more common. Therefore, researchers have examined how the routine activities of individuals affect their likelihood of various forms of victimization, including property crime (Mustaine & Tewksbury, 1998), violent crime (Sampson, 1987), and stalking (Fisher, Cullen, & Turner, 2002)
Miethe, Stafford, and Long (1987) argued that the routine activities of individuals differentially place some people and/or their property in the proximity of motivated offenders, thus leaving them vulnerable to victimization. Using victimization data from the 1975 National Crime Survey, Miethe et al. explored whether individuals’ major daily activities and frequency of nighttime activities affected their likelihood of property and violent victimization. Their analyses indicated that individuals who performed their major daily activities outside of the home had relatively higher risks of property victimization compared with those whose daily activities kept them at home. The location of major daily activities, however, was not significantly related to the risk of violent victimization. In terms of the frequency of nighttime activities, Miethe et al. found that individuals with a high frequency of nighttime activities were at an increased risk for property and violent victimization.
D. Routine Activities and Multilevel Opportunity
At first, researchers examined macro- and microlevel hypotheses derived from routine activities separately. Macrolevel routine activities have been used to explain crime rates, and the routine activities of individuals have been used to explain victimization risk. In more recent years, researchers have begun to explore whether opportunity factors operate at both individual and neighborhood levels to impact victimization risk (e.g., Sampson & Wooldredge, 1987; Wilcox Rountree, Land, & Miethe, 1994). In other words, do the routine activities of the neighborhood in which an individual resides independently influence his victimization risk beyond the effect of his own characteristics and routine activities that leave him vulnerable to crime? For example, leaving one’s door unlocked might contribute to victimization risk; living in a neighborhood where it is common to leave one’s door unlocked might also contribute to victimization risk. In the first case, one’s house can be easily entered if a burglar should try to enter. In the second case, a burglar knows to try to enter the home given the neighborhood norm of leaving doors unlocked. These two factors may both contribute to the risk of victimization for this individual.
In addition, researchers have questioned whether the effects of individual routine activities on victimization risk vary by neighborhood. For example, does leaving one’s door unlocked increase risk for burglary victimization to a particular level, regardless of whether one lives in the suburbs or in the city, or do the neighborhood characteristics condition the effect of individual routine activities on victimization risk? Routine activities theory and these types of research questions have inspired further theoretical developments in the area of multilevel opportunity (Wilcox, Land, & Hunt, 2003).
To answer these questions, researchers use data on both the characteristics of the neighborhood that indicate opportunities for crime, as well as the routine activities and other characteristics of the victim that might put him at risk for victimization. To analyze such data, researchers rely on sophisticated multilevel modeling techniques that allow them to determine the effects of individual- and neighborhood-level factors at the same time, as well as the extent to which neighborhood characteristics might condition the effects of individual routine activities on victimization risk.