U.S. Department of Justice Office of Justice Programs Bureau of Justice Statistics Crime Data Brief National Crime Victimization Survey Carjacking, 1993-2002 July 2004, NCJ 205123 ------------------------------------------------------------- This file is text only without graphic and many of the tables. A Zip archive of the tables in this report in spreadsheet format (.wk1) and the full report including tables and graphics in .pdf format are available from: http://www.ojp.usdoj.gov/bjs/abstract/c02.htm This report is one in a series. More recent editions may be available. To view a list of all in the series go to http://www.ojp.usdoj.gov/bjs/pubalp2.htm#carjacking -------------------------------------------------------------- By Patsy Klaus, BJS Statistician On average, 1993-2002, about 38,000 carjacking victimizations occurred annually, according to victim self-reports to the National Crime Victimization Survey (NCVS). There were an average 1.7 victimizations per 10,000 persons annually. About 15 murders a year involved car theft, though not all were carjack- ings, for the theft could have been incidental. About 34,000 carjacking incidents occurred annually, some having more than 1 victim. About 45% of these incidents were completed crimes. * Carjacking rates were higher on average during the first 5 years of the 1993-2002 period (2.1 per 10,000 persons each year) than during the last 5 years (1.3 per 10,000). * Men were more often victimized than women; blacks more than whites; and Hispanics more than non-Hispanics. * Households with annual incomes of $50,000 or more had lower rates than those making below $50,000. * Carjacking victimization rates were highest in urban areas, followed by suburban and rural areas. Ninety- three percent of carjackings occurred in cities or suburbs. * A weapon was used in 74% of carjacking victimizations. Firearms were used in 45% of carjackings, knives in 11%, and other weapons in 18%. * The victim resisted the offender in two-thirds of carjackings. Twenty- four percent of victims used confrontational resistance (threatening or attacking the offender or chasing or trying to capture the offender). About a third of victims used nonconfrontational methods, such as running away, calling for help, or trying to get the attention of others. * About 32% of victims of completed carjackings and about 17% of victims of attempted carjackings were injured. Serious injuries, such as gunshot or knife wounds, broken bones, or internal injuries occurred in about 9%. More minor injuries, such as bruises and chipped teeth, occurred in about 15%. * The victims were hospitalized in about 1% of carjackings. About 14% of victims were treated in hospital emergency departments and then released. Another 6% involved victims who were treated elsewhere, such as in a doctor's office or at the crime scene. About 4% of injured carjacking victims did not seek medical treatment. * Multiple carjackers committed about 56% of the total. * Males committed 93% of carjacking incidents, while groups involving both males and females committed 3%. Women committed about 3% of carjackings. * Carjacking victims identified 56% of the offenders as black, 21% as white, and 16% as members of other races, such as Asians or American Indians. In 6% of carjackings, the victim(s) reported multiple offenders of more than one race. * 68% of carjacking incidents occurred at night (6 p.m. - 6 a.m.). Forty-two percent of carjackings at night and 50% of those in the day were completed. * 44% of carjacking incidents occurred in an open area, such as on the street (other than immediately adjacent to the victim's own home or that of a friend or neighbor) or near public transportation (such as a bus, subway, or train station or an airport), and 24% occurred in parking lots or garages or near commercial places such as stores, gas stations, office buildings, restaurants/bars, or other commercial facilities. * About 63% of carjacking incidents occurred within 5 miles of the victim's home, including the 17% that occurred at or near the home. Four percent occurred more than 50 miles from the victim's home. * 77% of carjackings -- 98% of the completed crimes and 58% of the attempts -- were reported to the police. * Partial or complete recovery of property occurred in 78% of completed carjacking incidents. A quarter of carjackings involved total recovery of all property. According to the Supplemental Homicide Reports from the FBI's Uniform Crime Reports, about 15 murders a year involve auto theft. It is not possible to determine how many of these murders meet the definition of carjacking. ------------------------------------------------------------------- These data are from the National Crime Victimization Survey (NCVS), an ongoing survey which interviews about 42,000 households and 76,000 persons age 12 or older per year. Carjacking as defined in NCVS is a completed or attempted robbery in which a car or other motor vehicle was taken or an attempt was made to take it and the offender was a stranger to the victim. See for additional information on the NCVS methodology. Information about murder/auto thefts is from the Supplemental Homicide Reports of the FBI's Uniform Crime Reports. ----------------------------------------------------- Methodology Except for the information concerning murder/auto theft, this report presents data from the National Crime Victimization Survey (NCVS). This survey obtains information about criminal victimizations and incidents from an ongoing, nationally representative sample of households in the United States. NCVS data include both those incidents reported and those not reported to the police. Between 1993 and 2002 approximately 446,530 households and 831,060 persons age 12 or older were interviewed. Interviews were conducted with households and individuals twice per year. Response rates between 1993 and 2002 ranged between 92% to 96% of eligible households and 87% to 93% of individuals. Carjacking is not one of the crimes measured by the NCVS on a yearly basis because of the low number of cases identified annually (see further discussion below). To measure non-fatal carjacking over multiple years using NCVS data, cases were selected based upon the following criteria: (a) a completed or attempted robbery occurred, (b)a car or other motor vehicle, such as a truck or any other motorized vehicle legally allowed on public roads or highways, was taken or an attempt was made to take such a vehicle, and (c) the offender (or offenders) was a stranger to the victim(s). During the period of 1993-2002, a total of 130 unweighted cases of nonfatal carjacking were identified, 58 completed and 72 attempted. These cases were weighted to obtain national estimates of nonfatal carjacking incidents and carjacking victimizations in the United States. Aggregating data over a number of years allows examination of the characteristics of carjackings not possible when examining data for individual yeas. Although the number of sample cases for carjackings is too low to present yearly trend data, there are fewer unweighted cases of carjacking in recent years than in the early 1990's. If comparisons are made with the previous carjacking report, the declines in the average annual number of carjackings reflect the declining number of victimizations which have been occurring in recent years for most NCVS crimes, including carjacking. Carjackings can be analyzed as incidents (in which more than one victim may be present) or as victimizations (in which the characteristics of each victim are described). The terms "incidents" and "victimizations" are used to distinguish whether a characteristic refers to the entire incident (such as time or place of occurrence) or whether it describes the victim characteristics (such as age, race and gender). Information about murder/auto thefts was obtained from the Uniform Crime Reports sponsored by the Federal Bureau of Investigation. Data were obtained from Supplemental Homicide Reports covering the period of 1993 through 2002. Computation of standard errors The comparisons made in this report were tested to determine whether the observed differences between groups were statistically significant. Except where otherwise noted, comparisons mentioned in this report passed a hypothesis test at the .05 level of statistical significance (or the 95% confidence level). This means that the estimated difference between comparisons was greater than twice the standard error of that difference. For comparisons that were found to be statistically significant at the 90% confidence level, terms such as "somewhat more likely" were used. Calculations were conducted with statistical programs developed specifically for the NCVS by the U.S. Bureau of the Census. These programs take into account the complex NCVS sample design when calculating generalized variance estimates. Caution should be used when comparing estimates not discussed in the text. Seemingly large differences may not be statistically significant at the 95% of even the 90% confidence level. End of file 07/20/04 ih