Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
â2â Goals of the National Crime Victimization Survey I T IS EASY TO UNDERESTIMATE how little was known about crimes and victims before the ï¬ndings of the National Crime Victimization Survey (NCVS) became common wisdom. In the late 1960s, knowledge of crimes and their victims came largely from reports ï¬led by local police agen- cies as part of the Federal Bureau of Investigationâs (FBI) Uniform Crime Reporting (UCR) system, as well as from studies of the ï¬les held by individ- ual police departments. At the time, UCR coverage of the nation was far from complete andâin any eventâalmost all of the information it gathered came in highly aggregated form. Agencies sent in monthly totals for crimes in seven categories, which the FBI later published as yearly totals. The UCR program produced national data on the characteristics of speciï¬c incidents for only one crime type (homicide, through the Supplementary Homicide Reports). It produced no data on characteristics of victims, the costs or inju- rious consequences of crime, or the circumstances in which crimes occurred. Data on offenders came separately, in summary descriptions of the age, sex, and race of persons who were arrested, and national coverage of arrestees was even more incomplete than incident reporting. Furthermore, at that timeâbefore the dawn of the information technology revolutionâpolice recordkeeping was cumbersome and in many places haphazard, and it was widely feared that crime data were subject to manipulation if not misreport- ing, for political and organizational reasons. 25
26 SURVEYING VICTIMS Criminologists understood that there existed a âdark ï¬gureâ of crime consisting of events not reported to the police (Biderman and Reiss, 1967). Scattered studies of victimization, principally in Europe, suggested that the dark ï¬gure might be both very large and highly selective, raising important questions about what could be learned from crime data. Crime surveys promised to provide an alternative, often better, and cer- tainly more adaptable source of data on crimes and victims, one that could shine new light on the dark ï¬gure. Surveys commissioned by the Presidentâs Commission on Law Enforcement and Administration of Justice (1967) dur- ing the 1960s led toward the NCVS, which was further shaped by an exten- sive program of methodological research during the 1970s. As it developed, the survey proved to be a source of several important streams of new data, giving analysts more ï¬exibility in using them to address policy and research questions. In evaluating and assessing any program, it is essential to consider the programâs goals: what the program is intended to do and how objectives may have changed over time. Accordingly, this chapter reviews the fundamental goals of the NCVS, starting with a review of goal statements that have been advanced over the years (Section 2âA). In subsequent sections, we discuss some of the major themes in these goal statements in more detail: NCVS as a source of data on victims of crime (2âB), as counterpart to the UCR and a check on crimes reported to the police (2âC), as a tool for analytic ï¬exibility in studying different crime types (2âD), as a source for information on special topics in crime (2âE), and as a response to speciï¬c legal mandates (2âF). We return to discussion of speciï¬c goals at the close of Chapter 3, after surveying major issues facing the NCVS. 2âA HISTORICAL GOAL STATEMENTS OF THE NCVS As part of its comprehensive review, the Presidentâs Commission on Law Enforcement and Administration of Justice (1967:38) observed that âone of the most neglected subjects in the study of crimes is its victims: the persons, households, and businesses that bear the brunt of crime in the United States.â The commission argued: Both the part the victim can play in the criminal act and the part he could have played in preventing it are often overlooked. If it could be determined with sufï¬cient speciï¬city that people or businesses with cer- tain characteristics are more likely than others to be crime victims, and that crime is more likely to occur in some places than in others, efforts to control and prevent crime would be more productive. Then the pub- lic could be told where the risks of crime are greatest. Measures such as preventive police patrol and installation of burglar alarms and special locks could then be pursued more efï¬ciently and effectively. Individ-
GOALS OF THE NATIONAL CRIME VICTIMIZATION SURVEY 27 uals could then substitute objective estimation of risk for the general apprehension that today restrictsâperhaps unnecessarily and at best haphazardlyâtheir enjoyment of parks and their freedom of movement on the streets after dark. To get an initial reading of such information, the commission contracted with the National Opinion Research Center to conduct a National Survey of Criminal Victims. âThe ï¬rst of its kind conducted on such a scope,â the crime commissionâs survey reached a sample of 10,000 households. The survey results sufï¬ced to demonstrate to the commission that âfor the Na- tion as a whole there is far more crime than ever is reportedâ (Presidentâs Commission on Law Enforcement and Administration of Justice, 1967:v): Burglaries occur about three times more often than they are reported to police. Aggravated assaults and larcenies over $50 occur twice as often as they are reported. There are 50 percent more robberies than reported. In some areas, only one-tenth of the total number of certain kinds of crimes are reported to the police. Based on these ï¬ndings, the commissionâs ï¬nal report recommended the creation of a National Criminal Statistics Center, a âmajor responsibilityâ of which âwould be to examine the Commissionâs initial effort to develop a new yardstick to measure the extent of crime in our society as a supple- ment to the FBIâs Uniform Crime Reportsâ (Presidentâs Commission on Law Enforcement and Administration of Justice, 1967:x).1 That is, in addition to obtaining information from the oft-neglected perspective of victims of crime, the commission envisioned what would become the NCVS as a com- plementary measure to the crimes-reported-to-police measures of the UCR program. Notably, the commission emphasized this new data collection ve- hicle as a complement to, and not a replacement for, the UCR. Elsewhere in the report, the commission reiterated that âthe development of public sur- veys of victims of crime . . . can become a useful supplementary yardstickâ and emphasized (Presidentâs Commission on Law Enforcement and Admin- istration of Justice, 1967:31) that: what is needed to answer questions about the volume and trend of crime satisfactorily are a number of different crime indicators showing trends over a period of time to supplement the improved reporting by police agencies. Furthermore, the commission set forth a more ambitious goal for victimiza- tion surveys: providing information on the nature and levels of crime for 1 A previous National Commission on Law Observance and Enforcement (the Wickersham Commission) made a similar call for criminal statistics program administered by an indepen- dent, central statistical agency; however, that recommendation languished (Cantor and Lynch, 2000:95).
28 SURVEYING VICTIMS areas smaller than the nation as a whole. âThe Commission believes that the Government should be able to plot the levels of different kinds of crime in a city or a State as precisely as the Labor Department and the Census Bureau now plot the rate of unemployment. Just as unemployment information is essential to sound economic planning, so some day may criminal infor- mation help ofï¬cial planning in the system of criminal justiceâ (Presidentâs Commission on Law Enforcement and Administration of Justice, 1967:x). As described in Box 1-1, the crime commissionâs report led to discus- sions between the U.S. Department of Justice and the Census Bureau on implementing a national survey of victims. An initial planning conference held by the Census Bureau articulated nine possible goal statements for the new data collection program (U.S. Bureau of the Census, 1968): ⢠Provide an independent calibration for the UCR; ⢠Provide a measure of victim risk; ⢠Enable a shift in concentration in the criminal justice system from the offender to the victim; ⢠Provide an indicator of the crime problem outside those indicators generated by police activity; ⢠Serve as an index of changes in reporting behavior in the population; ⢠Provide an indicator of the social âoutlookâ in the population as well as an indicator of societyâs deï¬nitions of crimes; ⢠Serve as a basis for the study of granting of compensation to victims; ⢠Serve as a statistic to determine the degree of involvement by the vic- tim; and ⢠Serve as a measure of public conï¬dence in police. In 1970, the ï¬rst proposed study plan for a national victimization survey reï¬ned this list and suggested that the surveyâs primary purpose would be âto measure the annual change in crime incidents for a limited set of major crimes and to characterize some of the socio-economic aspects of both the reported events and their victimsâ (reprinted in Work, 1975:220). An unattributed document titled âObjectives of the National Crime Surveyââdated âcirca 1976â2 and apparently an internal Law Enforcement 2 The documentâprovided by BJS to the panel as part of a brieï¬ng package for its ï¬rst meetingârefers brieï¬y to âthe Academy reportâ and questions that âthe Academy raised,â re- ferring to Surveying Crime (National Research Council, 1976b). However, it also includes reference to the commercial surveys that were part of the original National Crime Surveys but were dropped on the basis a recommendation in National Research Council (1976b). Hence, it appears to postdate the National Research Council report but not by a great deal. The language in this document is different fromâand more speculative thanâthe formal statements of LEAA
GOALS OF THE NATIONAL CRIME VICTIMIZATION SURVEY 29 Assistance Administration (LEAA) documentâis broad in scope. It declares that âthere is only one unchanging objectiveâ of the survey but describes that objective in a manner that is different from past versions: âto learn about crime and criminal justice and closely related phenomena, that which can only (or best) be learned from a large representative National survey.â This statement emphasizes the victimization surveyâs methodological inde- pendence from the UCR and ofï¬cial reports to police. Indeed, the document resists âthe conceptual subjection of the NCS to the UCRâ: âthe ability of the NCS to supplement, complement and enhance the UCR does exist,â but the surveyâs inherent capability for âmuch broader and more varied approaches to crime and its consequencesâ is inherently limited âto the extent the NCS limits its conceptualization to the UCR framework.â Accordingly, âit is not an objective of the NCS to measure simply what the UCR does not measure, nor to measure âbetterâ what the UCR does measure.â The LEAA document notes three corollaries to this âunchanging objectiveâ: 1) to continuously pursue a program of conceptual and methodolog- ical research on a scale necessary to support such a large scale research effort and to maximize its capacity to inform and en- lighten; 2) to fully exploit the capacity of the survey to inform and enlighten through a program of competent, creative, and thorough substan- tive analytical research; 3) to disseminate research ï¬ndings to researchers, practitioners, pol- icymakers and the general citizenry in forms most appropriate to insuring that the ï¬ndings become part of the working knowledge of each audience. The document also speciï¬ed that certain statements should not be goals of the victimization survey. Although âgood NCS data for metropolitan ar- easâ might have âunique and valuable research potential,â the document argues that âthe NCS is not a suitable mechanism for informing local law enforcement planners and administrators about crime within their jurisdic- tion.â Likewise, the survey âcannot be, and must not be promoted as being, a vehicle for evaluating local programs aimed at reducing crime or improv- ing the criminal justice system.â Following the publication of Surveying Crime (National Research Coun- cil, 1976b), the Law Enforcement Assistance Administration (1981) listed seven formal âobjectives of a publicly funded nationwide statistical series on victimization,â emphasizing that the objectives are âcurrentâ and hence subject to revision: objectives issued in 1975 (Work, 1975) and 1977 (Law Enforcement Assistance Administration, 1981), before and after the National Research Council report.
30 SURVEYING VICTIMS ⢠To provide trend data that will serve as a set of continuous and comparable national social indicators for the rate of victimization for selected crimes of violence and crimes of theft and for other factors related to crime and victimization in support of national criminal justice policy and decisionmaking and in support of in- formed public discussion. ⢠To conduct a program of conceptual and methodological research that will improve the victimization surveys in response to the Na- tional Academy of Sciences evaluation, including reï¬nements of measurement, survey techniques, and questionnaire design. ⢠To exploit the depth and richness of currently available victimiza- tion data through analytical research on issues of public concern and consequence to the development of national, State, and local criminal justice policy and legislation, with broad dissemination of ï¬ndings. ⢠To assist State and local government efforts to improve the ad- ministration of criminal justice through (a) promotion of analysis of national data to understand local implications; (b) provision of national guidance on the feasibility, conduct, and utility of local victimization surveys; and (c) provision of a limited set of subna- tional social indicators derived from the national survey. ⢠To expand the current victimization survey to include assessment of vulnerability and susceptibility to crime of various segments of the population, and to explore governmental and private ap- proaches for reducing the opportunity for criminal acts and the risk of victimization. ⢠To examine, through the longitudinal component of the survey, those factors associated with repeated or multiple victimizations to discover appropriate means of reducing such victimizations or minimizing their consequences. ⢠To use the ongoing national survey to obtain additional informa- tion on crime and criminal justice issues through supplemental questionnaires. The consortium tasked with redesigning the National Crime Survey in the mid-1980s reasoned that âthe process of constructing an efï¬cient sample design begins with a detailed description of survey objectives and a clear ranking of these objectives.â Hence, it developed its own ranked list of objectives (highest to lowest) and considered costs and beneï¬ts of proposed changes relative to each of these objectives (Biderman et al., 1986:19â20):
GOALS OF THE NATIONAL CRIME VICTIMIZATION SURVEY 31 Recommended Ranking of NCS Objectives 1.0 To provide data that will serve as a set of continuous and com- parable social indicators of trends in the rates of victimization by selected crimes of violence and crimes of theft and of factors re- lated to crime and victimization in support of national criminal justice policy, and informed public decision. 2.0 To provide policymakers at the national and state and local levels as well as the research community, with a database concerning crime victims and victimization. 2.1 To provide empirical information concerning the characteris- tics of victims and consequences of the victimization that will be useful in designing, implementing, and maintaining victim assistance programs. 2.2 To provide empirical information on perceived satisfaction with the criminal justice system. 3.0 To facilitate analytical research on issues of public concern and of consequence to the development of national, state, and local criminal justice policy. 3.1 To provide empirical information that assists individuals and households in avoiding victimization. 3.2 To provide empirical information relevant to understanding the difference between the reported crime rate (UCR) and the victimization rate. 4.0 To provide, in addition to national data, crime indicators for se- lected cities and states. 4.1 To assist state and local governments in evaluating the feasi- bility and utility of local victimization surveys. 5.0 To gather information on a regular basis concerning attitudes to- ward crime, criminals, and crime control. Most of theseânotably excluding the provision of âcrime indicators for se- lected cities and statesââwere retained in a 1989 listing of âcurrent BJS objectives for the NCS programâ (Bureau of Justice Statistics, 1989:5â6). Current documentation for the public-use NCVS ï¬les archived by the Interuniversity Consortium for Political and Social Research (Bureau of Jus- tice Statistics, 2007a:5) provides a concise summary of historical goal state- ments: âThe NCS and NCVS were both designed with four primary objec- tives: 1) to develop detailed information about the victims and consequences of crime; 2) to estimate the numbers and types of crimes not reported to po- lice; 3) to provide uniform measures of selected types of crimes; and 4) to permit comparisons over time and types of areas.â Likewise, the cur- rent Census Bureau manual for NCVS interviewers (U.S. Census Bureau, 2003:A1-2) simpliï¬es the discussion to a âprimary purposeâ and a âsec- ondary purposeâ:
32 SURVEYING VICTIMS [The primary purpose is] to obtain, from respondents who are 12 years of age and older, an accurate and up-to-date measure of the amount and kinds of crime committed during a speciï¬c six-month reference period.3 . . . The NCVS also collects detailed information about speciï¬c incidents of criminal victimization that the respondent reports for the six-month reference period. [The secondary purpose is that] NCVS also serves as a vehicle for obtaining supplemental data on crime and the criminal justice system; . . . this supplemental information is collected periodically, along with the standard NCVS data. 2âB DATA ON CRIMES AND VICTIMS At the most fundamental level, the NCVS is intended to do what the name implies: provide measures of criminal victimization. This statement is a great understatement, though, as the implementation of the NCVS opened up several dimensions of understanding of crime that were not well or sys- tematically considered in earlier measures: ⢠Unreported crime: Before the ï¬rst crime surveys, no one had an inkling of the magnitude of the âdark ï¬gureâ of unreported crime, nor was there any systematic knowledge of the factors that facilitated or dis- couraged victim reporting. The NCVS examines whether victimiza- tions were reported to police and the reasons for reporting or not reporting. See Section 3âF for further discussion. ⢠Characteristics of victims: Before the NCVS, little was known about crime victims except for sketchy proï¬les (typically of their age, sex, and race) that could be extracted with some difï¬culty from police case ï¬les. There was great interest in learning more about them. At the time the NCVS was being developed, victims were frequently regarded as the âforgotten participantsâ in the criminal justice system. There was concern at the time about apparently high rates of victimization of the elderly and about domestic violence. The survey promised to shed new light on the victims of different kinds of crime, including a range of victim attributes (such as household income or education) that go un- recorded by the police. See, for example, Klaus and Rennison (2002) and Bachman et al. (1998) on differential patterns in victimization by age; Smith (1987), Smith and Kuchta (1993), Lauritsen and Schaum (2004) on victimization of women; Dugan and Apel (2003) on com- bined effects of race, ethnicity, and gender; and Levitt (1999) and Thacher (2004) on difference in victimization by income. 3 The wording of this objective is somewhat odd, as it suggests to interviewers that the goal of the survey is to elicit confessions (i.e., âcrimes committedâ) from respondents.
GOALS OF THE NATIONAL CRIME VICTIMIZATION SURVEY 33 ⢠Consequences for victims: Detailed questions on the NCVS yielded the ï¬rst national picture of the consequences of crime for victims, in- cluding extent of injury, hospital stays and insurance coverage, out- of-pocket medical expenses, the value of lost time from work, and other costs. The resulting national estimates have been used in cost- beneï¬t analyses of crime prevention programs (for a review, see Co- hen, 2000), and they make the NCVS the only comprehensive na- tional source of information on the costs of crime to victims (see Sec- tion 3âE.1). ⢠Circumstances surrounding victimization: The survey includes ques- tions about the location of each incident (at home, in commercial places, at school, etc.); whether household members or other people were present at the scene and if they took any action; and if respon- dents took any self-protective measures that appear to have improved their situation. Other details include gun and weapon use, whether the crime was an attempted or completed incident, and the number of offenders. See, for example, Rand (1994) and Zawitz (1995) for NCVS ï¬ndings regarding the use of ï¬rearms in crime and Bachman et al. (2002) on the use and effective of self-protective behaviors by victims. ⢠Proï¬les of offenders: In surveys, victims can be asked to describe of- fenders and what could be discerned concerning their motivations. For example, among those victims who provided information about the offenderâs use of alcohol, about 30 percent of the victimizations in- volved an offender who had been drinking (http://www.ojp.usdoj.gov/ bjs/cvict_c.htm; Greenfeld and Henneberg, 2001). The NCVS also in- cludes questions about the role of apparent bias in the targeting of victims, to probe the prevalence of hate crimes (as we discuss further in Section 2âF). 2âC NATIONAL BENCHMARK: NCVS AND THE UNIFORM CRIME REPORTS Along with the basic goal of providing information on the characteristics of victims, the role of measurements from a victimization survey as a coun- terpart to ofï¬cial reports to policeâshedding light on the so-called dark ï¬gure of unreported crimeâis an original, long-standing theme. The role of NCVS as a âcontinuous and comparable national social indicationâ to the UCR programâs ofï¬cial counts is directly envisioned by the BJS authoriz- ing legislation, as discussed in Section 2âF. The duality between the NCVS and the UCR is a major part of the publicâs awareness of the victimization survey: each year, the release of new data from the NCVS and UCR pro-
34 SURVEYING VICTIMS grams is met with a ï¬urry of news stories on possible upticks and downticks in crime. But the duality may also contribute to confusion and uncertainty among the public and stakeholders like Congress, when duality and com- plementarity are mistaken for redundancy and it is assumed that the two programs measure exactly the same thing. The conundrum is that the po- tential for confusion is heightened on those occasions when the NCVS and the UCR suggest trends that go in opposite directions, while the potential for perceived redundancy and waste is heightened when the two measures suggest similar trends. Given this chapterâs focus on historical goals of the NCVS, this section deals with only a limited set of issues: basic deï¬nitional differences between the two programs and the national role of the NCVS as a counterpart in- dicator to the UCR results. Chapter 3 examines broader dimensions of the duality between the NCVS and the UCR, including the tracking of trends in the two series over time (convergence and divergence relative to each other) and the relative quality of the series. Technical details about the UCR are included for reference in Appendix D. When UCR collection began in 1929, the program was designed to col- lect information âon a subset of crimes that the International Association of Chiefs of Police (IACP) at the time considered prevalent, serious, and well reported to policeâ (Cantor and Lynch, 2000:86â87). That core set of crime typesâthe so-called index crimes of criminal homicide, forcible rape, aggra- vated assault, robbery, burglary, larceny-theft, and motor vehicle theftâhas remained the same over time. Indeed, as detailed in Appendix D, only the crime of arson has subsequently been elevated to the UCRâs top tier of mea- sured offenses. As a survey, the NCVS offered (and still does) a basis for collecting in- formation on more and varied types of victimization incidents than police records permit. Yet the NCVS remains decidedly tethered to some features and concepts of the longer established UCR program. Cantor and Lynch (2000:107) argue that the initial design of the National Crime Survey de- viated from the recommendations of the Presidentâs Commission on Law Enforcement and Administration of Justice (1967) that led to its founding due to âthe attempt to mimic UCRâ: Speciï¬cally, the principle of facilitating recall and reporting was com- promised somewhat in favor of some of the legal principles and the desire to classify crimes neatly. . . . [The new NCS, administered by the Census Bureau,] restricted its screening to Part I crimes in UCR, such that questions were asked with the intent of eliciting mentions of these crimes and only these crimes. Although the Census instrumentation separated the screening task from the provision of detailed information for classiï¬cation, there was a one-to-one correspondence between the screen questions and the UCR crimes.
GOALS OF THE NATIONAL CRIME VICTIMIZATION SURVEY 35 Although the NCVS may have echoed UCR structures in order to estab- lish an equal footing, it remains true that the NCVS is designed to produce that which the UCR âhas never attempted to produce: a count of crime that includes serious offenses, like rape, that may never be reported to po- liceâ (Robinson, 2007). In providing these data âin a reliable and consistent fashion,â Robinson (2007) argues that BJSâ sponsorship of the NCVS ï¬lls âa distinctly federal role,â generating data that no single state can afford to produce on a regular basis.4 We discuss victimization surveys that have been conducted by individual states in Section 3âD. 2âD ANALYTIC FLEXIBILITY In the early 1970s, as now, the UCR gathered monthly and yearly crime totals in only a few (currently eight) broad categories, and the FBI received them only at the jurisdiction level. One promise of the NCVS was that the incident-level data it produced could be used to address a variety of research and policy questions, because of the analytic ï¬exibility of surveys. One fea- ture of this ï¬exibility is that the NCVS can be analyzed at multiple levels. The survey gathers reports of individual and household victimization, and most descriptive publications examine rates of crime at those levels. How- ever, the data can be organized in a variety of ways to address descriptive and analytic questions. ⢠Households âtouched by crimeâ: BJS reports have combined data for households and all of the individuals living in them, to characterize the percentage of households that have had some recent experience with crime (e.g., Klaus, 2007). Families are another analytic unit that can be distinguished in the survey, and a variety of family crimes are within the scope of the survey (Durose et al., 2005). ⢠Crimes by location: Fairly detailed descriptions of the location of inci- dents are gathered in the NCVS, revealing that 22 percent of the vic- tims of violence were involved in some form of leisure activity away from home at the time of their victimization; 22 percent reported they were at home, and another 20 percent mentioned they were at work or traveling to or from work when the crime occurred. See, for exam- ple, Warchol (1998) for speciï¬c study of workplace violence. Schools are another important locus for crime problems, and the survey has been used extensively to examine school crime (as described further in the next section). 4 Robinson (2007), testifying before a congressional appropriations subcommittee, argued that NCVS should be afforded âa broadened role in helping in our understanding of victim- ization. BJS should be provided with increased funding to enable it to measure crime on a state-by-state basis, and even to the level of large cities.â
36 SURVEYING VICTIMS ⢠Neighborhood effects on victimization: For special analyses, data for the census tracts in which NCVS respondents live have been combined with their responses to the survey, to examine the effects of such fac- tors as concentrated poverty and residential instability on victimization rates and the characteristics of crimes (Baumer, 2002; Baumer et al., 2003). ⢠Metropolitan area rates and trends: The sample design of the NCVS is capable of providing estimates of crime for the largest metropolitan areas. NCVS data for some of the largest areas have been used to describe local area victimization trends and to compare these to trends in the UCR for the same areas (Lauritsen and Schaum, 2005; Lauritsen, 2006a). ⢠Long-term trends: With the exception of the redesign in 1992, the methodology of the NCVS has remained fairly consistent over time. The data have been used to track long-term national trends in crime as well as long-term victimization trends for different subgroups, such as males and females, blacks and whites, and different age groups (see, e.g., Steffensmeier and Harer, 1999; Klaus and Rennison, 2002). ⢠Longitudinal patterns: Because NCVS respondents are interviewed up to seven times while they are participants in the crime panel, data can be organized to trace patterns of victimization over time for in- dividuals and households that do not change residences. These data have been used to study such issues as the effect of reporting crime to the police on subsequent victimization and the effect of victimization on the decision to move (see, e.g., Conaway and Lohr, 1994; Dugan, 1999). 2âE TOPICAL FLEXIBILITY: NCVS SUPPLEMENTS From the beginning it was anticipated that the NCVS would provide a ï¬exible vehicle for gathering occasional or one-time data to supplement the ongoing core data required to track national trends in crime. These supple- ments might be supported by research grants, allocations by Congress, or by contributions by partner agencies who wished to gather specialized data relevant to their responsibilities. One responsibility is to design ways to fold these requests into ongoing data collection in ways that do not disrupt the continuing ï¬ow of data. The supplements have made signiï¬cant contribu- tions to research and policy. They include: ⢠Crime seriousness: The ï¬rst NCVS supplement gathered national data on the perceived seriousness of crime, information that has been used to differentially weight incidents to reï¬ect their impact on the pub-
GOALS OF THE NATIONAL CRIME VICTIMIZATION SURVEY 37 lic. Subsequent BJS analyses have used both NCVS data and infor- mation from the American Housing Survey to describe perceptions of crime severity (DeFrances and Smith, 1994); on a one-shot basis, BJS also collaborated with the Ofï¬ce of Community Oriented Policing Ser- vices in conducting community safety surveys by telephone in 12 cities, wholly distinct from the NCVS (Smith et al., 1999). ⢠Attitudes and lifestyles: Another supplement gathered extensive data on the attitudes of individuals and the relationship between crime and how they conduct their lives (Murphy, 1976; Cowan et al., 1984). ⢠Crime in schools: BJS, in collaboration with the National Center for Education Statistics, periodically collects data on aspects of school crime through the School Crime Supplement to the NCVS. Respon- dents age 12 and older attending school are asked about their school environment. Information is gathered on the availability of drugs at school, the existence of street gangs and the prevalence of gang ï¬ghts, the presence of guns at school, victimizations, and their fear of be- ing attacked or harmed. Results from the various administrations of the School Crime Supplement, as well as related data resources, are described by Dinkes et al. (2007). Another important supplement, the Police-Public Contact Survey, is de- scribed in the next section as an example of a legislative mandate. 2âF LEGAL MANDATES Finally, in outlining the goals and objectives of the NCVS, it is important to consider responsibilities that the survey (and BJS) must accomplish due to requirements in law. The duties and functions of BJS laid out in its enabling statute [42 U.S.C. 3732(c)(2)â(3), originally enacted as P 90-351, the Omnibus Crime Con- .L. trol and Safe Streets Act of 1968] directly authorizes the bureau to âcollect and analyze information concerning criminal victimization, including crimes against the elderly, and civil disputes.â The authorizing language further di- rects that BJS should: collect and analyze data that will serve as a continuous and comparable national social indication of the prevalence, incidence, rates, extent, distribution, and attributes of crime, juvenile delinquency, civil disputes, and other statistical factors related to crime, civil disputes, and juvenile delinquency, in support of national, State, and local justice policy and decisionmaking. Although the use of the resulting data for developing national policy is noted in the language, the preamble to BJSâ authorization clearly hearkens to the
38 SURVEYING VICTIMS agencyâs origins as part of the Law Enforcement Assistance Administration: in its work, BJS âshall give primary emphasis to the problems of State and local justice systemsâ (42 U.S.C. 3731). Over the years, Congress has mandated that speciï¬c information related to criminal victimization be collected by BJS. Some of these mandates ex- plicitly direct that these new data collections be added to the NCVS, while others become linked to the NCVS because the insertion of a supplement to the survey is seen as the most expedient solution. An example of a direct-to- NCVS mandate is a provision in P 106-534, the Protecting Seniors from .L. Fraud Act of 2000. As part of a broader study of the prevalence of crimes against seniors, section 6 of the act requires BJS, âas part of each National Crime Victimization Survey,â to: include statistics relating toâ (1) crimes targeting or disproportionately affecting seniors; (2) crime risk factors for seniors, including the times and locations at which crimes victimizing seniors are most likely to occur; and (3) speciï¬c characteristics of the victims of crimes who are seniors, including age, gender, race or ethnicity, and socioeconomic status. Response to this legislative mandate has focused on the crimes of identity theft, credit card fraud, and bank fraud. Similar language in P 105-301 directed that the NCVS include mea- .L. sures of âthe nature of crimes against individuals with developmental dis- abilitiesâ and âthe speciï¬c characteristics of the victims of those crimes.â This act, the Crime Victims with Disabilities Awareness Act of 1998, led to implementation (after pilot testing) of NCVS questions gauging whether vic- tims of crime were in poor health, had any physical or mental impairments, or had disabilities that affected their everyday life. They are also asked to judge if any of these provided an opportunity for their victimization. The Violent Crime Control and Law Enforcement Act of 1994 (P 103- .L. 322, Section 210402) included a brief provision that âthe Attorney Gen- eral shall, through appropriate means, acquire data about the use of exces- sive force by law enforcement ofï¬cers.â As a direct result of this mandate, BJS developed the Police-Public Contact Survey to measure the extent of all types of interactions between the police and members of the public (of which those involving âexcessive forceâ is logically a subset). The survey was ï¬rst conducted on a pilot basis in 1996 (Greenfeld et al., 1997); after reï¬ne- ment, it was fully ï¬elded as an NCVS supplement in 1999 and has become a continuing occasional supplement. The survey gathers detailed information about the nature of police-citizen contacts, respondent reports of police use of force and their assessments of that force, and self-reports of provocative actions that they may have themselves initiated during the encounter. Data
GOALS OF THE NATIONAL CRIME VICTIMIZATION SURVEY 39 from the supplement have been used to in analyses of possible levels of racial proï¬ling by, for example, Engel and Calnon (2004) and Engel (2005). Other examples of legislative mandates impacting the NCVS include: ⢠The Hate Crime Statistics Act of 1990 (P 101-275), which directed .L. the Attorney General to âacquire data, for each calendar year, about crimes that manifest evidence of prejudice based on race, religion, dis- ability, sexual orientation, or ethnicity.â In this case, the NCVS was not explicitly mentioned as the data collection vehicle. In response to the mandate, the NCVS now includes questions probing the pos- sible role of prejudice or bigotry in motivating offenders. For results from the hate crime questions on the NCVS, see Harlow (2005), and see Lee et al. (1999) Lauritsen (2005) for additional discussion of the development of the questions. We discuss the addition of hate crime questions further in Section 3âC.1. ⢠Family violence reporting provisions in P 100-690, which directed .L. BJS, âthrough the annual National Crime Survey, [to] collect and pub- lish data that more accurately measures the extent of domestic violence in America, especially the physical and sexual abuse of children and the elderly.â The addition of NCVS content and topics due to legislative mandates implicitly recognizes the importance of the NCVS as a source of informa- tion. However, it also adds to the burden on the survey and on BJS as a whole, particularly since these changes have not been matched by budgetary increases.