The Justice Department announced
yesterday that violent crime fell 15 percent last year, the largest drop
on record, a finding that puzzled some criminologists because it came
only two weeks after the Federal Bureau of Investigation reported that
serious crime remained stable in 2000, ending an eight-year period of
significant declines in crime.
The new report, based on a survey of 160,000 crime victims
nationwide, also found that rates of violent and property crime were at
their lowest levels since the Justice Department started keeping the
figures in 1973.
Criminologists expressed confusion yesterday at the discrepancy
between the two reports, both released by the Justice Department,
although they said the differences could be partly explained by how the
reports measure crime. The new report, called the National Crime
Victimization Survey, uses a sampling technique to measure how many
Americans are victims of crime in a year. The F.B.I. report, known as
the Uniform Crime Report, is compiled from crimes reported to police
departments across the nation.
The F.B.I. report measures only the most serious crimes, including
homicide, rape, robbery, aggravated assault, burglary, car theft and
larceny. The Crime Victimization Survey does not count homicide, because
it questions crime victims, but it does include simple assaults, like
pushing and shoving matches, which are more numerous than aggravated
assaults or robberies and therefore tend to statistically dominate the
Simple assaults accounted for 61.5 percent of all violent crime in
the new survey, and because they declined by 14.4 percent in 2000
compared with 1999, they accounted for most of the large drop in violent
crime, said James Alan Fox, the Lipman family professor of criminal
justice at Northeastern University and an adviser to the Justice
Department on crime data.
In the F.B.I. report, published on May 30, violent crime was said to
have increased by 0.1 percent in 2000 compared with 1999.
A factor adding to the experts' uncertainty over the new report is
that for most of the 1990's as crime fell, the F.B.I.'s statistics and
those of the victimization survey closely mirrored each other. That
provided confidence that the crime drop was real. But the divergence in
the latest two reports raises a question of whether one of them has gone
Lawrence Greenfeld, the acting director of the Bureau of Justice
Statistics, said yesterday that the differences between the two
estimates were not as great as they might appear. Part of the reason,
Mr. Greenfeld said, is that simple assaults, including domestic
violence, make up a large part of the victims' survey but are not
counted by the F.B.I.
Another factor, he said, is that the largest part of the decrease in
violent crimes in the victims' survey came from crimes not reported to
So which report more accurately measures crime? It depends on what a
person considers important, Professor Fox said. For most Americans, who
worry about homicide, rape and robbery when they think about crime, the
F.B.I. report provides a more meaningful portrait of what is happening,
he said, and it showed that after dropping since 1992, crime leveled off
"I wouldn't emphasize the victimization report too much because
it measures more the less serious crimes," he added.
But Alfred Blumstein, a leading criminologist at Carnegie Mellon
University, said the real story behind the difference in the two reports
is, "They are telling us that crime is very difficult to
Both reports have their strengths and weaknesses, Professor Blumstein
said. "It's like the Florida election results," he said.
"These are issues that are difficult to sort out. I wouldn't just
accept one and reject the other. We have to live with this cloud of
The F.B.I. report can be misleading because it depends on reports of
crimes to the police and then on police forces turning in their data to
the bureau. Generally, only 80 percent of the nation's police forces
provide their figures, Professor Blumstein said. The victimization
survey was started in 1973, at the suggestion of criminologists, to
provide a check on the police figures.
One of its benefits is that it provides much more information on the
circumstances of a crime than the F.B.I. figures. For example, the
victimization survey includes information on the location of the crime,
the weapon used and what the relationship was, if any, between the
victim and the criminal.
"I think that crime did level off last year," Professor
Blumstein said. "But the victimization report raises questions
about that." To be sure that crime did stop falling, he said,
experts must wait for more than one year's figures.
To compile the victimization report for the Justice Department, the
Census Bureau interviewed 160,000 people 12 years and older in a
nationally representative sample.
Over all, the victimization report found that violent crime fell 15
percent last year, while property crime dropped 10 percent.
Almost every group in the population experienced less violent crime
last year, the report said, including men, women, whites, blacks and 12-