Implications

NIBRS data from the jurisdictions reporting in 1997 suggest that the current practice of differentiating the crime of kidnaping into only two categories (family and nonfamily) needs to be changed. The two components of the conventional nonfamily category—acquaintance kidnaping and stranger kidnaping—seem to be different types of offenses, at least as they appear in law enforcement data. The specific characteristics of acquaintance kidnaping, which has not yet been separately profiled, need to be better delineated and understood.

Acquaintance kidnaping in data from NIBRS jurisdictions distinguishes itself from stranger kidnaping in a variety of important ways (table 1). First, it involves more juvenile offenders and somewhat more female offenders. Second, it occurs more often with teenage victims, while more stranger kidnaping victimizes school-age children (the complement of "teenage victim"). Third, acquaintance kidnaping is much more likely to occur at a home or residence, while stranger kidnaping most often occurs in outdoor locales. Finally, acquaintance kidnaping victims suffer a higher rate of injury. These substantial differences highlight that acquaintance kidnaping is both a separate and serious form of kidnaping.

Unfortunately, because of the limited categories of information available in NIBRS, it is impossible to draw confident conclusions about the dynamics and motives that might specially characterize acquaintance kidnaping. Nevertheless, NIBRS data are consistent with case material suggesting that certain specific types of crimes are encompassed within the acquaintance kidnaping category. For example, one specific type of acquaintance kidnaping is the situation where boyfriends or former boyfriends kidnap girlfriends (32 percent of the acquaintance kidnaping of teenage female victims) to seek revenge for being spurned, force a reconciliation, commit a sexual assault, or perhaps evade parents who want to break up the relationship. Another type of acquaintance kidnaping is related to gang activity: for example, the situation where teenagers abduct other teenagers in order to intimidate, recruit, or retaliate against them. A third type of acquaintance kidnaping involves family friends or employees (for example, babysitters) who remove children from their home for the purpose of sexual assault or perhaps retaliation against the family. Such variety of offenders and motives may make acquaintance kidnaping more difficult to typify than family or stranger kidnaping, but it is nevertheless instructive in that it highlights the importance of obtaining data on a larger number of cases for the purposes of profiling.

There is no easy explanation for why acquaintance kidnaping involves more injury than stranger kidnaping. It may be the combination of more teenage victims who would be inclined to resist and the fact that intimidation is a more common motive for these crimes, which results in the use of more force and thus more injury. It may also be that police are less likely to think of kidnaping as an element in an acquaintance crime (and thus less likely to record kidnaping as an additional offense in the NIBRS database) unless the victim is injured.

It is also instructive to examine the discrepancies between the pictures of nonfamily kidnaping presented in NIBRS data and those found in the data collected by the earlier NISMART, which conducted an independent review of police files (abduction, missing persons, homicide, and sexual assault files) in 1988 (Finkelhor, Hotaling, and Sedlak, 1990). NISMART data record fewer acquaintance and juvenile perpetrators but substantially more weapon usage. Another significant discrepancy is in the percentage of kidnaping incidents associated explicitly with sexual assault. In NISMART, sexual assault appeared to be a motive in two-thirds of the nonfamily abductions known to police (Asdigian, Finkelhor, and Hotaling, 1995), whereas in NIBRS, only 15 percent of non-family kidnaping (of both male and female victims) was coded with the additional crime of sexual assault.

There are several possible explanations for these differences. The methodology used in NISMART to look for abductions was to directly sample police departments' sexual assault files but not their general assault or robbery files, which may have exaggerated the portion of kidnaping that was associated with sexual assault. At the same time, NISMART's direct access to police files may have revealed sexual assault motives or intentions that are not captured when officers apply NIBRS codes. It is certainly surprising that so much of the stranger and acquaintance kidnaping recorded in NIBRS has no other crime associated with it, because these types of kidnaping are generally considered as primarily a method to facilitate other offenses. One problem might be that law enforcement officials, continuing the UCR tradition with its single-code limit, do not take advantage of the multiple crime codes allowed under NIBRS and fail to record adjunct or secondary offenses. For whatever reason, NIBRS may underrepresent the motive of sexual assault and the presence of other crimes in acquaintance and stranger kidnaping.

Another conspicuous discrepancy between NIBRS and NISMART data concerns the relative occurrence of family and nonfamily abduction. The NISMART data suggest that the overwhelming majority of abductions were committed by family members (Finkelhor et al., 1990), whereas family abductions actually constitute only a slight minority of all kidnapings recorded in NIBRS. However, the NISMART family abduction estimates come from a national household survey, not police records, and, although 44 percent of the families surveyed indicated they had contacted police, relatively little is known about how such reports are recorded or tabulated in crime statistics. The discrepancy between the NISMART and NIBRS data suggests that many, if not most, calls to police about family abductions are not recorded as crimes.

Findings on kidnaping from the NIBRS jurisdictions also are inconsistent with other kidnaping studies—for example, those based on FBI data or national inquiries of police investigators (Boudreaux, Lord, and Dutra, 1999; Hanfland, Keppel, and Weis, 1997)—concerning characteristics such as the identity of perpetrators or the incidence of serious injury and death. These differences can almost all be traced to the different samples selected in different studies—for example, samples of kidnaping homicides. What is really highlighted by all of these comparisons is the absence of a consensus about how segments of the population of kidnaped children should be aggregated or subdivided for most useful policy analysis. NIBRS data provide yet another, but by no means a complete or conclusive, perspective on the problem.

The inconclusiveness of findings about kidnaping point to the main policy needs in this area. First, substantially more research about this crime—which has attracted a large amount of public attention but rather little scientific study or profiling—is needed. Second, research about the problem would benefit if those studying and collecting data about it would adopt a common set of definitions and categories within which to subdivide and analyze it. This uniformity has been achieved in regard to other crimes (e.g., sexual assault) in recent years as a result of national data systems like the UCR, but kidnaping was outside the purview of this system. The inclusion of kidnaping in NIBRS offers the opportunity to achieve that uniformity now. Third, in the light of this and in order to increase the usefulness of NIBRS for the study of kidnaping, NIBRS may need to make a special effort to help local agencies report kidnaping in uniform and consistent ways, since it is a crime that may be handled in disparate fashions from jurisdiction to jurisdiction.