Throughout the country as a whole, crime tends to be significantly underreported for a variety of reasons. According to a survey conducted by the Justice Center, this is true for Anchorage as well.
This article, the second in a series of Alaska Justice Forum articles on findings from the Anchorage Adult Criminal Victimization Survey (AACVS), will look at actual criminal victimization experiences of the survey respondents, including whether such victimizations were reported to the police. Also presented are findings from a national victimization survey and a similar local victimization survey conducted in 12 cities.
The Anchorage survey was administered between April 1 and June 30, 2002. Eligible respondents, who were residents age 18 or older, were contacted via a household (non-business) telephone line. In general, the telephone calls were made on weekdays between 10:00 AM and 9:00 PM, although calls were usually not made during the dinner hours between 5:00 PM and 7:00 PM. A random-digit dialing (RDD) method that generates phone numbers using a computer program was used. Since each household with a telephone had an equal chance of being contacted, this method increased the likelihood that the residents surveyed were, in fact, representative of Anchorage residents.
Interviewers explained the purpose of the study to potential respondents, guaranteed confidentiality, and asked for participation. It should be noted that although households were randomly called, no random selection of individuals within households occurred. While such randomization was attempted in the first few days of survey administration, interviewers realized that the number of callbacks necessary to secure an interview with a randomly selected respondent would be both time and cost prohibitive. Participating respondents within a household were selected simply based on who was willing to answer the survey questions (in most cases this was the individual answering the telephone). The overall survey cooperation rate (number of completed interviews divided by the sum of completed interviews, refusals, terminations, hearing/language problems, and lack of respondent availability) was approximately 60 percent, based on a total of 781 secured interviews.
As shown in Table 1, a comparison of AACVS respondent characteristics and Anchorage Census 2000 data shows strong similarities, but two key differences are worth noting. First, survey respondents were disproportionately female. Second, a smaller proportion of AACVS respondents reported household incomes of $50,000 or more, although this difference is likely due to the larger number of respondents who refused to answer the income question.
Much of the survey instrument addressed the victimization experiences, if any, that the respondent had suffered in the previous twelve months and the characteristics of those victimizations. Crimes covered by the survy include rape, other sexual assault, aggravated and simple assault, robbery, larceny, motor vehicle theft, and burglary. Respondents were never directly asked whether they had been the victim of a specific crime. Instead, specific crime labels were attached to incidents based on responses to a series of questions representing definitions of particular crimes. This more indirect approach helped to avoid situations where respondents were unaware of the definitions of crimes (e.g., distinction between a burglary and a robbery) and erroneously reported victimizations.
As Table 2 shows, the 781 participating respondents reported 284 separate incidents consistent with the definitions of crimes captured by the AACVS. These victimizations were experienced by 208 survey respondents—more than a quarter (26.6%) of the survey sample. Over 80 percent of the incidents were property crimes; 19 percent were personal/violent crimes; and the remaining one percent fell into the category of completed pocket picking. By far, the most common type of victimization was larceny or attempted larceny (n=185), accounting for nearly two-thirds (65.1%) of all victimizations reported. Larceny involves the attempted or actual theft of property or cash from the respondent or respondent’s household. Unlike robbery, larceny does not involve the use of force or threat of force and, unlike both robbery and pocket picking, it does not involve any type of personal contact.
The most common personal victimization was assault (n=28), including 12 aggravated assaults and 16 simple assaults. Aggravated assaults include attacks resulting in serious injury to the victim (e.g., weapon wounds, broken bones, loss of consciousness, internal injuries) or attacks, attempted attacks, or threatened attacks involving the use of a weapon, regardless of whether or not serious injury occurs. In contrast, simple assaults include attacks or attempted attacks without a weapon that result in minor injury (e.g., bruises, scratches, black eye, and cuts) or no injury to the victim. In addition to aggravated and simple assaults, threats of assault (other than those involving a weapon), rape, or sexual assault (n=19) were also relatively common.
Table 2 also includes rates for the various crimes, but, particularly for violent crimes, caution should be exercised in interpreting rates. Due to the small number of violent victimizations reported in the survey, violent crime rates are unstable—that is, they can shift dramatically as a result of small changes in the number of victimizations. This issue is less problematic with property victimizations, where the larger number of reported victimizations results in less sensitive rates of property victimization. Nevertheless, all rates are viewed best as rates of victimization for the survey sample rather than as estimates of the rate of victimization in Anchorage as a whole.
Table 3 shows, with greater specificity, the incidence of particular victimizations. This increased precision in definition demonstrates that the broad categories presented in Table 2 comprise varied incident types. Several examples illustrate such distinctions. First, the most common type of assault is one where no weapon was used and no injury occurred (46% of all completed assaults). Nevertheless, one-third (32.1%) of all assaults result in some type of injury to the victim. Second—as would be expected—most larcenies are completed larcenies (96.8%); respondents are not generally aware of a theft unless some item was taken. Finally, completed burglaries where no forcible entry was used (69.6%) were more common than completed burglaries when forcible entry was used (30.4%), perhaps reflecting the relative ease of accomplishing the former over the latter.
ersonal victimization rates for survey participants were higher for respondents who were male, Alaska Native, young, and/or from households with lower combined incomes. Table 4 shows the demographic characteristics of the victims in the 54 personal crimes. Again—many of the cells have very small counts, resulting in sensitive estimates.
Respondents experiencing property victimization most often reported that the crime occurred at or near their homes. In fact, one-third (67.5%) of all property crimes occurred in the respondent’s home or another structure on the property, in the apartment hall, storage area, or laundry room, in the yard or on the sidewalk, or on the street immediately adjacent to the home. Slightly more than one-third (37.0%) of respondents experiencing personal victimization indicated that the crime took place in or around their home. Over 15 percent (16.7%) of personal victimizations occurred on the street not adjacent to their own residence or the residence of a neighbor, friend, or relative. Other personal victimizations occurred in restaurants, nightclubs, or bars (7.4%) or inside some other commercial establishment (7.4%).
Reporting to Police
Victimization data are able to lend insight into crimes regardless of whether they are reported to the police. This is important because in Anchorage 63 percent of personal victimizations and 66.2 percent of property victimizations went unreported. Table 5 identifies the reasons given by respondents for not reporting the crime to the police. The most commonly cited reasons for not reporting personal crimes were that the incident was a private matter (23.5%) or it was reported to another official (23.5%). For property crimes, more than 40 percent (43%) indicated that the minor nature of the crime was a contributing factor in the decision not to report the crime to police. Other more practical considerations were given as additional reasons why property crimes were not reported: property could not be recovered (15.2%) or lack of proof (13.2%). In nearly 1 in 10 property victimizations where the incident was unreported, residents viewed the crime as unimportant to the police (9.3%).
Other Victimization Studies
In this section, findings from the Anchorage survey are presented in conjunction with findings from two victimization studies employing virtually identical survey instruments, but using slightly different methodologies. As discussed later, it is important to use caution in making comparisons. Nevertheless, placing the Anchorage surveys findings within a comparative context is worthwhile since the two comparison studies actually served as models for the AACVS.The AACVS resulted, in part, from a Bureau of Justice Statistics initiative encouraging states and local communities to conduct crime victimization surveys. This effort was designed to overcome one of the key limitations of the National Crime Victimization Survey (NCVS), which has been around since 1973. The NCVS is most useful in painting a national (or other large aggregate) picture of crime. Since it is based on a national sample of respondents, individual communities or states represent only a small portion of the overall sample, thereby prohibiting the extraction of reliable local (small area) crime statistics. In the late 1990s, the Bureau of Justice Statistics conducted a demonstration project designed to show the feasibility and utility of conducting local victimization surveys. This study, referred to here as the 12-City Study, led to production of a software program for administering local victimization studies. (It is described in the BJS report Criminal Victimization and Perceptions of Community in 12 Cities, 1998.)
Although the instrument used in each study was largely the same, there were several key differences in methodological approaches. First, the NCVS and the 12-City Study administer the survey instrument to all households members age 12 years and older, while the Anchorage survey is an adult (18 and older) victimization survey. This primarily presents a problem for violent crime rates, because each respondent identifies only personal victimization experiences. For property crimes, a single household respondent, regardless of age, acts as the informant for collecting information on all property crimes experienced by the household. Second, the AACVS and 12-City Study used telephone surveys, while the NCVS still collects much of its data through person-to-person interviews. Third, NCVS respondents are interviewed multiple times over a three-year period, thereby reducing reporting errors associated with recall. In contrast, the AACVS and the 12-City Study participants were interviewed only once. Finally, each of the three studies was administered during a different time period.
Table 6 presents data from all three studies—the Anchorage study, the 12-City Study, and the 2001 NCVS (the most recent year for which published data are available). The findings—crime rates and reporting rates—from the second two studies are presented only to show what has been found in other cities and in the nation as a whole. Due to the diverse methodologies used in each study, no firm conclusions can be drawn from comparisons; that is, we cannot say definitely that property crime rates are greater in Anchorage than in Los Angeles, for example.The caveats aside, Table 6 does show that findings from Anchorage are generally consistent, both in victimization rates and reporting rates, to findings from the 12-City Study, the study most similar in methodology to the AACVS. The NCVS results shown are lower. Unlike the Anchorage study, which sampled a non-rural area, the national figures derive from a sample of urban, suburban and rural areas. The NCVS urban-only victimization rates for 2001 (not shown) were 33.2 and 212.8 for personal and property victimizations, respectively.
Other findings from the AACVS will appear in future issues of the Alaska Justice Forum. Some of the topics to appear in forthcoming articles include respondents’ perceptions of the police as well as AACVS findings disaggregated by the geographic location of the respondents’ residences.
The complete results of this analysis will be available in a final report in fall 2003. Matthew Giblin is Assistant Professor of Criminal Justice at York College of Pennsylvania. From 2000 to 2002, he was a research associate with the Justice Center.