Evaluating Recent Changes in Homicide and Robbery Rates

• Author: Brian E. Oliver,Richard Rosenfeld
• DOI: 10.3818/JRP.10.2.2008.49
• Date: 01 December 2008
• Published date: 01 December 2008
• Subject Matter: Research Note



*



Evaluating Recent Changes in Homicide and

Robbery Rates

Richard Rosenfeld

Brian E. Oliver

Department of Criminology and Criminal Justice

University of Missouri-St. Louis

* Abstract

This paper evaluates the Police Executive Research Forum’s (PERF) recent warnings of

a “gathering storm” of criminal violence in the United States. We argue that increases

in violent crime during 2005 and 2006 were the expected result of year-to-year changes

in the economy and imprisonment. Our argument is supported by an analysis of

changes in homicide and robbery rates between 1970 and 2006 showing that short-run

variation in homicide and robbery rates are a function of changes in imprisonment, the

age composition of the population, and the economy. But PERF’s warning of a new

crime wave may simply have been premature. The steep economic downturn of 2008

portends a crime rise that may dwarf the 2005 and 2006 increases. Local communities

and criminal justice practitioners should demand better research on crime trends that

can help them evaluate current practice and plan for the future.

JUSTICE RESEARCH AND POLICY, Vol. 10, No. 2, 2008

© 2008 Justice Research and Statistics Association

Research Note



In August of 2006 the Police Executive Research Forum (PERF) held a “crime sum-

mit” in Washington, D.C., to warn the nation of an impending increase in violent

crime. The meeting was attended by police chiefs and other off‌icials from cities around

the country and resulted in a report, entitled A Gathering Storm: Violent Crime in

America, documenting violent crime increases in those jurisdictions during 2005 and

2006 (PERF, 2006). The meeting and subsequent PERF report generated national pub-

licity (Willing, 2006) and calls for the federal government to assist cities in combating

the crime rise, especially by hiring more police. The Department of Justice responded

by conducting its own investigation of local crime problems (Johnson, 2006), and

in

2007 the Bureau of Justice Assistance awarded nearly $75 million to over 100 law

enforcement agencies to assist in local efforts to combat violent crime (see http://www.

ojp.usdoj.gov/BJA/grant/tvc.html). In the spring of 2007, PERF published another re-

port containing updated crime f‌igures and a renewed call to action (PERF, 2007a).

This paper evaluates the claim of a “gathering storm” of violent crime in the

United States. We argue that increases in violent crime during 2005 and 2006 did

not diverge markedly from expectations based on prior f‌luctuations associated with

changing economic conditions and reduced growth in imprisonment. This argu-

ment is supported by the results of f‌ixed-effects panel models of robbery and ho-

micide rates between 1970 and 2006, as well as by preliminary Uniform Crime

Reports (UCR) data for 2007 showing decreases in both homicide and robbery

rates. It appears that the storm clouds broke, but only temporarily. Early signs point

to another crime rise that may dwarf those of 2005 and 2006 and for which local

communities should begin to plan ahead.

* A Gathering Storm?

PERF based its claim of an upswing in violent crime on data furnished by more

than 50 urban police departments represented at the 2006 crime summit. Although

the departments were to some degree self-selected and therefore the national rep-

resentativeness of the PERF sample of jurisdictions is questionable, UCR crime

f‌igures for 2005 and 2006 later conf‌irmed nationwide increases in violent crime.

Homicide and robbery rates rose 1.8% and 3.0%, respectively, between 2004 and

2005. Homicide rose another 1.8% and robbery increased 6.1% in 2006.1 PERF

regarded a two-year increase in violent crime as a portentous trend reversal that,

1

See Table 1 of the 2006 Uniform Crime Reports (http://www.fbi.gov/ucr/cius2006/

data/table_01.html). The current analysis of violent crime focuses on homicide and robbery

and omits aggravated assaults and rapes. Rapes are notoriously underreported in the UCR

and evidence indicates that UCR aggravated assault trends are biased because of changes over

time in the classif‌ication and recording of assaults by the police (see Rosenfeld, 2007).