On the epidemiology of violent juvenile crime in America: a total arrest-referenced approach.

The catalyst for this study was a widely publicized U.S. government-sponsored report forecasting alarming increases in violent juvenile crime. Working from data for the entire United States given in annual FBI Uniform Crime Reports, the study presented a descriptive statistical, historical profile of violent juvenile crime in America based on the percentage of all arrests for criminal homicide, forcible rape, robbery, and aggravated assault attributable to juveniles, 1941-1995. Markedly different patterns resulted from focusing on rates and rates of change calculated for 1-, 5-, and 10-yr. periods. Thus, for example, the most recent 5-yr. trend for criminal homicide indicated an average annual increase of 7.81% in the incidence of arrests attributable to juveniles, with a corresponding projected increase of 117% from 1995 to 2010. An average annual rate increase of 5.13% was indicated over the most recent 10-yr. period, leading to a predicted increase of 77% in 2010 as compared to 1995. By sharp contrast, focusing on the 15.3% rate increase which occurred in the most recent single year led to the expectation that juvenile arrests will account for 229% more of this nation's criminal homicide arrests in 2015 than was the case in 1995. In every case, widely discrepant, 15-yr. projection differences such as those noted above, are magnified considerably if we assume validity of U.S. census estimates concerning increases in the size of America's juvenile population over the time period considered. Results of the study are taken to underscore the importance of qualifying archival data-based inferences about violent juvenile crime, in terms of the specific measure(s) used and time-frame context(s) of the unit(s) of analysis.