Crime May Not Be As Bad As The Numbers Make It Look...

By all accounts, violence is spinning out of control in the U. S., imposing a huge and growing economic burden on the nation. Since 1980, for example, the American prison population has doubled to more than 1 million, and the bill for correctional systems has tripled to well over $22 billion.

In an article in the latest issue of The American Prospect, however, Northwestern University sociologist Christopher Jencks contends that fears of an ever-escalating wave of violence are exaggerated. While many factors lie behind such fears, Jencks believes they have been inflamed by a methodological distortion in the crime statistics issued by the Federal Bureau of Investigation.

No one denies that violent crime rates--crime per capita--at least doubled from 1963 to 1974. But since then, annual FBI estimates of violent crime (mainly murder, robbery, and aggravated assault) have tended to move ever higher, whereas reports from the Bureau of Justice Statistics (BJS) have almost always shown the crime rate as constant or declining. The BJS data are based on national victimization surveys of the public, whereas the FBI statistics are based on crimes reported by police departments to the agency.

Why the discrepancy? When the first crime victimization surveys were conducted in the early 1970s, notes Jencks, it became clear that police records reflected only a small proportion of reported violent crimes. So the Justice Dept. began working with local police departments to improve their record keeping.

As a result of these continuing efforts, writes Jencks, police reports to the FBI have included crimes that tended to be unrecorded in the past, imparting an upward bias to the FBI crime count. By contrast, the BJS victimization data show no clear trend in violent crimes from 1973 to 1981, an actual decline in crime rates during most of the 1980s, and an apparent pickup in recent yearsbut with violence still below the levels recorded in the 1970s.

Jencks also notes that though murder rates rose in New York, Detroit, and Washington in the 1980s, they declined in Chicago, Los Angeles, and Houston. Murder rates among blacks fell and then rose a bit in the 1980s but at last count were still lower than in the 1970s. Rates of aggravated assault and robbery committed by blacks dropped by 25% to 30% from 1973 to 1987, whereas similar crimes by whites declined less sharply.

As for demographics, Jencks estimates that the level of violent crime should have risen about 8% from 1960 to 1975 and then fallen by 7% from 1976 to 1988 if it were determined solely by the percentage of males in the high-crime 15-to-24-year age group. In actuality, it doubled in the first period and then declined by about 15% in the second. In short, says Jencks, "it's premature to conclude that violence is in a long-term upward trend."

Before it's here, it's on the Bloomberg Terminal.