oversight

Information Generally Not Available on Toy Gun Issues Related to Crime, Injuries or Deaths, and Long-Term Impact

Published by the Government Accountability Office on 2003-09-30.

Below is a raw (and likely hideous) rendition of the original report. (PDF)

United States General Accounting Office
Washington, DC 20548



          September 30, 2003

          The Honorable Edolphus Towns
          House of Representatives

          Subject: Information Generally Not Available on Toy Gun Issues Related to Crime,
          Injuries or Deaths, and Long-Term Impact

          Dear Mr. Towns:

          This report responds to your request that we provide you with information on several
          issues related to the use of toy guns. Specifically, you asked that we (1) examine
          crime statistics showing the prevalence of crimes that involved toy guns in some
          capacity; (2) gather any available information on incidents involving toy guns that
          have resulted in injuries or deaths, whether or not related to criminal activity; and
          (3) determine from available literature whether there are any studies examining the
          long-term impacts that can be attributed to toy gun play by children.

          As agreed with your office, we focused our study on imitation or look-alike toy guns
          and excluded toy guns that fire projectiles, for example, BB guns, paintball guns, and
          pellet guns. To obtain relevant information, we conducted an extensive literature
          search using the Internet and other electronic resources to identify applicable
          statistics, reports, studies, articles, or other publications. In addition, we contacted
          federal officials at various agencies, including the Department of Justice’s Bureau of
          Justice Statistics (BJS), the Department of Health and Human Services’ Centers for
          Disease Control and Prevention (CDC), and the Consumer Product Safety
          Commission (CPSC). Furthermore, we interviewed university researchers or
          academicians and contacted the counsel for the Toy Industry Association. We
          performed our work from May to August 2003 in accordance with generally accepted
          government auditing standards. Enclosure I presents more details on our objectives,
          scope, and methodology.

          Results in Brief

          Our study disclosed that scant data exist on the incidence of crimes, injuries, or
          deaths involving toy guns and on the long-term effects that childhood play with toy
          guns may have on individuals. Available data on crimes involving toy guns are dated
          and insufficient for providing a national perspective. Also, databases that collect
          information from hospital emergency rooms and other sources regarding product-
          related injuries and deaths generally are not designed to capture information about
          incidents involving toy guns. Thus, the relatively few cases of such incidents that
          were recorded in these databases probably do not represent an accurate or


                                                                       GAO-03-1135R Use of Toy Guns
comprehensive reporting. Finally, our literature search found no publications or
studies specifically addressing the long-term effects of childhood play with toy guns.

Crimes Involving the Use of Toy Guns

In response to our inquiries, officials at three of the Department of Justice
components we contacted—the Bureau of Alcohol, Tobacco, Firearms and
Explosives; the National Institute of Justice; and the Federal Bureau of
Investigation—said they had no information about crimes involving the use of toy
guns. Generally, the only data we found regarding the use of toy guns in crimes are
presented in a June 1990 report—Toy Guns: Involvement in Crime & Encounters
With Police—prepared by the Police Executive Research Forum (PERF), under a
cooperative agreement with BJS.1 In response to our inquiry, in June 2003, BJS
informed us the agency has no current plans to sponsor or undertake a follow-up
study to update the 1990 report.

In conducting its study, PERF surveyed 699 state and local law enforcement agencies
and received 458 usable responses (a response rate of 66 percent). PERF’s report
does not include information on whether the nonrespondents differed in significant
ways from the respondents. Without such information, it is not possible to determine
if the lack of response from 34 percent of the agencies distorted the findings.

Among other questions, the survey solicited information on the number of robberies
and assaults that involved the use of toy guns during the period January 1, 1985, to
September 1, 1989. According to PERF, police department reporting systems typically
are not coded to identify the involvement of imitation or toy guns in crimes. As a
result, most responding agencies provided information from either a manual records
check or a solicitation of information from officers. Relying on officers’ memories
may have resulted in either an under- or over-reporting of incidents involving toy
guns. For the period January 1, 1985, to September 1, 1989,

• 	 148 law enforcement agencies (32 percent of the 458 usable responses) reported a
    total of 2,796 robberies committed with the use of toy guns and

• 	 121 law enforcement agencies (26 percent of the 458 usable responses) reported a
    total of 3,104 assaults committed with the use of toy guns.

As a collateral issue, PERF also reported that—for the period January 1, 1985, to
September 1, 1989—law enforcement agencies seized a total of 10,065 toy guns.
According to PERF, this total does not include guns that were stolen property.
Rather, the total consists only of those toy guns that were directly or indirectly
involved in an incident—such as robbery, assault, domestic disturbance, suspicious
person, etc.—where the police took some form of action.


1
 PERF is a national membership organization of police executives from the largest city, county, and state law enforcement
agencies.



Page 2                                                                                  GAO-03-1135R Use of Toy Guns
The PERF researchers concluded that insufficient data were available to clearly
determine whether the use of toy guns to commit crimes was a serious problem,
particularly in comparison to all crimes of violence and police-involved shootings
throughout the nation. As noted above, two factors—response rate issues and
concerns about the reliability of information based to some extent on officers' recall
of incidents—result in reservations about the findings. Enclosure II presents more
details about PERF’s report.

Injuries or Deaths Involving Toy Guns

One way an injury or death could occur is for a police officer to mistake a toy gun for
a real firearm. As part of PERF’s survey, researchers asked law enforcement agencies
to report the number of incidents where officers had used actual force (deadly or less
than deadly) based on the belief that a toy gun was real. For the study period (Jan. 1,
1985, to Sep. 1, 1989), 31 law enforcement agencies (7 percent of the 458 usable
responses) reported a total of 105 applicable incidents where officers had used actual
force, either deadly or less than deadly (see table 3 in enc. II). PERF’s report did not
specify how many of the 105 incidents resulted in injuries nor how many resulted in
deaths.

To further determine the availability of information on incidents involving toy guns
that have resulted in injuries or deaths, whether or not related to criminal activity, we
contacted two federal agencies—CPSC and CDC—that have databases with
information on health and/or safety issues. At our request, CPSC officials reviewed
the agency’s three major databases that provide information on product-related
hazards, and the officials reported the following to us:

• 	 National Electronic Injury Surveillance System. This system collects
    information and provides national estimates on the number of victims treated in
    hospital emergency rooms for product-related injuries. The entire system includes
    98 hospitals reporting almost 700,000 cases each year. For the period January
    2000 to July 2003, the database showed 301 incidents of injuries that resulted from
    many hazard patterns involving toy guns, including children swinging or throwing
    toy guns. According to CPSC, the 301 incidents encompass the broad category of
    toy guns—not just the limited category of replica guns. CPSC’s review of about
    100 of these incidents that involved victims aged 10 years or older disclosed no
    injury incidents in which police officers mistook toy guns for real firearms.

• 	 Death Certificate System. This database contains records of death certificates
    for product-related deaths from each of the 50 states. The database excludes data
    on firearm-related deaths but may record deaths involving toy guns not caused by
    a firearm. For the most recent 10 years (1993 through 2002), the database showed
    no cases of deaths involving toy guns or situations where police officers mistook
    toy guns for real firearms.




Page 3                                                        GAO-03-1135R Use of Toy Guns
• 	 Incident Data Base. This data file contains records of cases received from news
    clips, medical examiner reports, consumer complaints, and reports from other
    sources. For the most recent 10 years (1993 through 2002), the data file showed
    four cases of individuals being fatally shot by police officers who mistook toy
    guns for real firearms.

CPSC officials told us that these statistics on police shootings involving toy guns
likely represent an undercounting of such incidents. The officials explained that
CPSC would not generally collect data on police shootings because the agency’s
focus is on consumer product issues rather than firearms.

CDC is a data contributor and has inquiry access to CPSC’s National Electronic Injury
Surveillance System. At our request, CDC officials reviewed the system’s database.
For the 1 year reviewed (2001), the officials reported identifying 66 incidents
involving toy guns. Of these 66 incidents, the majority (62) involved individuals 0 to
19 years of age, and the remaining 4 incidents occurred among the over-19 age group.

According to the CDC officials, of the 62 incidents among individuals 0 to 19 years
old, 57 incidents involved unintentional injuries, such as choking on a toy gun part or
being hit with toy gun or projectile part. The other 5 incidents apparently were
reported to the police as being assaults and involved children hitting other children
with a toy gun while fighting or engaging in rough play. CDC officials described these
5 incidents as follows:

•    A 13-year old child was hit on head with a plastic gun.
•    A 12-year old female was hit on the elbow by a toy gun.
•    A 12-year old male was injured in a fight involving a toy gun.
•    A 7-year old was poked in the right eye with a plastic gun.
•    An individual was struck on the head by a toy gun.

Also, CDC officials told us that the agency’s National Violent Death Reporting System
(NVDRS)—being designed to collect information on all violent deaths, including
those involving toy guns—was not yet operational. The officials referred us to
Harvard University’s National Violent Injury Statistics System (NVISS), which is a
pilot program for CDC’s NVDRS and encompasses 12 sites nationwide.2 The co­
director of the pilot program told us that the NVISS database contained information
for 2 years (2000 and 2001) but does not include a variable to facilitate an electronic
search for injuries or deaths involving toy gun incidents. Nevertheless, the co-director
responded to our questions based on her knowledge of the database and her review
of the more recent year’s (2001) data for 7 sites—Connecticut, Maine, Wisconsin,
Utah, San Francisco, Miami-Dade County, and Allegany County. For the 2001 data,
the co-director reported finding no deaths involving toy guns. In addition, the


2
 The 12 sites consist of 6 states (Connecticut, Kentucky, Maine, Maryland, Wisconsin, and Utah); 

5 metropolitan areas or counties—Atlanta (Ga.), Detroit (Mich.), Miami-Dade County (Fla.), Allegany County (Pa.), and 

San Francisco (Cal.)—and 1 pilot (conducted by the University of Pennsylvania) collectively covering three metropolitan 

areas—Bethlehem (Pa.), Youngstown (Ohio), and Iowa City (Iowa).




Page 4                                                                                  GAO-03-1135R Use of Toy Guns
co-director did not recall seeing any toy gun-related deaths in the first year’s (2000)
data.

Our literature search found that, in 1987, the American Academy of Pediatrics’
Committee on Accident and Poison Prevention issued a policy statement, which said
that, “The main hazard presented by nonprojectile toy guns is that children who play
with them may inadvertently be drawn to playing with real weapons which they
mistake for toys.”3 The policy statement recommended that pediatricians counsel
parents concerning the hazards of having toy guns in the house.

Long-Term Effects of Childhood Play with Toy Guns

Our literature search of social science, scientific, educational, crime and justice, and
other journals and publications disclosed no authoritative study on the possible long-
term effects on individuals of childhood play with toy guns. Generally, the literature
discussed numerous possible causes of aggressive behavior, including exposure to
violence in video games and television, and did not focus specifically on childhood
play with toy guns.

One exception we found was a 1992 Brandeis University study, entitled “The Relation
Between Toy Gun Play and Children’s Aggressive Behavior.”4 The study was based on
a small number of preschoolers in one daycare center and found limited evidence
that toy gun play was associated with increased real aggression and with decreased
pretend aggression in free-play settings. However, due to the small number of
children involved, all from one location and in one setting, the results are not
generalizable to other children. Further, the analytical method used may have
overstated the significance of the association between toy gun play and aggression. In
addition, the study did not examine whether longer-term associations between toy
gun play and future aggression are likely.

Furthermore, we contacted the senior scientific editor of a 2001 report—Youth
Violence: A Report of the Surgeon General—to discuss the extent, if any, that the
study addressed or considered the long-term impacts attributable to toy gun play by
children.5 This individual said that the Surgeon General’s study focused on violence
involving real firearms and did not consider toy gun issues. He expressed
unawareness of any research on the long-term effects of childhood play with toy
guns.

Agency Comments

We provided a draft of this report for comment to the Department of Justice, the
Department of Health and Human Services, and CPSC. During the period September
17-25, 2003, we received written or oral comments from these agencies. The

3
  American Academy of Pediatrics, “Injuries Related to ‘Toy’ Firearms” (RE7085), Pediatrics, Volume 79, Number 3 (Mar. 1987).

4
  Malcolm W. Watson and Ying Peng, Brandeis University, “The Relation Between Toy Gun Play and Children’s Aggressive

Behavior,” Early Education and Development, Volume 3, Number 4 (Oct. 1992).

5
  Youth Violence: A Report of the Surgeon General, U.S. Public Health Service (Washington, D.C.: Jan. 2001).




Page 5                                                                                  GAO-03-1135R Use of Toy Guns
comments indicated that the draft accurately presented the information that we had
obtained from the respective agency. Also, the Department of Health and Human
Services (including CDC) and CPSC provided technical clarifications, which we
incorporated in this report where appropriate.
                                      _____

As agreed with your office, unless you publicly announce its contents earlier, we plan 

no further distribution of this report until 30 days after its issue date. At that time, we 

will provide copies to the Attorney General, the Secretary of Health and Human 

Services, and the Executive Director of CPSC. We will also make copies available to 

others on request. 


If you have any questions about this report, please contact me at (202) 512-8777 or 

Assistant Director, Danny R. Burton, at (214) 777-5600. Other key contributors to this 

report were Fredrick D. Berry, Ann H. Finley, SaraAnn W. Moessbauer, 

Julia A. Rachiele, Miguel A. Salas, and Susan B. Wallace. 


Sincerely yours, 





Laurie E. Ekstrand 

Director, Homeland Security and Justice Issues 


Enclosures-2 





Page 6                                                          GAO-03-1135R Use of Toy Guns
Enclosure I

                       Objectives, Scope, and Methodology

Objectives

Representative Edolphus Towns asked us to (1) examine crime statistics showing the
prevalence of crimes that involved toy guns in some capacity; (2) gather any available
information on incidents involving toy guns that have resulted in injuries or deaths,
whether or not related to criminal activity; and (3) determine from available literature
whether there are any studies examining the long-term impacts that can be attributed
to toy gun play by children.

Scope and Methodology

As agreed with the requester’s office, we focused our study on imitation or look-alike
toy guns and excluded toy guns that fire projectiles, for example, BB guns, paintball
guns, and pellet guns. To obtain relevant information, we conducted a literature
search using the Internet and other electronic resources to identify applicable
statistics, reports, studies, articles, or other publications. Specifically, we used
keywords/key phrases to search the following three major sources:

• 	 Dialog. Dialog provides access to over 800 databases covering scientific and
    technical literature, trade journals, and newswires. Our search covered the period
    January 1, 1990, to July 10, 2003.

• 	 Nexis. Nexis provides access to news stories in major U.S. newspapers. Our
    search covered the period January 1, 2000, to July 11, 2003.

• 	 Nexis’ Statistical Universe. This source provides access to three data files—(1)
    the American Statistics Index, which covers statistical publications of the U.S.
    government; (2) the Statistical Reference Index, which covers statistical
    publications from sources other than the U.S. government; and (3) the Index to
    International Statistics, which covers international publications. Our search of
    these three data files covered the periods beginning in 1973, 1980, and 1983,
    respectively, to July 31, 2003.

In addition to the literature search, we contacted various federal agencies, university
researchers or academicians, and a representative of the Toy Industry Association
(see table 1).




Page 7                                                       GAO-03-1135R Use of Toy Guns
Enclosure I


Table 1: Federal Agencies, Universities, and the Trade Association GAO Contacted


    Federal agencies                                              Position of person contacted
    Bureau of Alcohol, Tobacco, Firearms and                      Audit Liaison
    Explosives
    Bureau of Justice Statistics                                  Statistical Policy Advisor for Bureau of Justice
                                                                  Statistics
    Centers for Disease Control and Prevention—                   Director, Office of Statistics and Programming
    National Center for Injury Prevention and Control             Deputy Director; Office of Policy, Evaluation, and
                                                                  Legislation
                                                                  Medical Epidemiologist, Division of Violence
                                                                  Prevention
    Consumer Product Safety Commission                            Executive Director
                                                                  Deputy Executive Director
                                                                  Assistant Executive Director for Hazard Identification
                                                                  and Reduction
                                                                  Program Manager, Hazard Identification and
                                                                  Reduction
                                                                  Director, Data Systems - Epidemiology
    Federal Bureau of Investigation                               Attorney Liaison
    National Institute of Justice                                 Audit Liaison, Office of Justice Programs
                  a
    Universities
    University of California, Davis (Davis, Cal.)                 Director, Violence Prevention Research Program
                                                                  (Sacramento, Cal.)
    University of Central Florida (Orlando, Fla.)                 Provost Distinguished Research Professor,
                                                                  Department of Sociology and Anthropology
    University of Colorado (Boulder, Colo.)                       Director, Center for the Study and Prevention of
                                                                                                            b
                                                                  Violence, Institute of Behavioral Science
    Harvard University (Boston, Mass.)                            Co-Director, National Violent Injury Statistics System
    University of Georgia (Athens, Ga.)                           Associate Professor, Department of Health Promotion
                                                                  and Behavior
    New York University School of Law (New York, N.Y.)            Director, Center for Research in Crime and Justice
    Trade association
    Toy Industry Association (New York, N.Y.)                     Counsel
Source: GAO.
a
 Our contacts with universities were based on suggestions made by federal agency officials and the results of our literature
search.
b
This university official served as the senior scientific editor for the Surgeon General’s 2001 report on youth violence.


Crime Statistics Involving Toy Guns

To determine the availability of statistics regarding the prevalence of crimes that
involved toy guns in some capacity, we contacted the following Department of
Justice components:

• 	 Bureau of Justice Statistics (BJS). We followed-up on the results of our
    literature search, which identified a June 1990 report (Toy Guns: Involvement in
    Crime & Encounters with Police) prepared by the Police Executive Research
    Forum, under a cooperative agreement with BJS (see enc. II).




Page 8                                                                                    GAO-03-1135R Use of Toy Guns
Enclosure I


• 	 National Institute of Justice (NIJ). NIJ is the research, development, and
    evaluation agency of the Department of Justice.

• 	 Federal Bureau of Investigation (FBI). We inquired whether the Uniform
    Crime Reporting Program collected any information involving toy gun incidents.

• 	 Bureau of Alcohol, Tobacco, Firearms and Explosives (ATF). We inquired
    whether the agency’s “Youth Crime Interdiction Initiative” collected any
    information involving toy gun incidents.

Toy Gun Incidents Resulting in Injuries or Deaths

To determine the availability of information on incidents involving toy guns that have
resulted in injuries or deaths, whether or not related to criminal activity, we
contacted two federal agencies—the Consumer Product Safety Commission (CPSC)
and the Department of Health and Human Services’ Centers for Disease Control and
Prevention (CDC). These agencies collect data on health and safety issues in several
databases.

At our request, CPSC officials reviewed the agency’s three major surveillance
databases that provide information on product-related hazards:

• 	 National Electronic Injury Surveillance System. This system collects
    information and provides national estimates on the number of victims treated in
    hospital emergency rooms for product-related injuries. From this database, CPSC
    officials reviewed cases from January 2000 to July 2003.

• 	 Death Certificate System. This database contains records of death certificates
    for product-related deaths from each of the 50 states. CPSC officials reviewed
    cases reported during the previous 10 years (1993 through 2002).

• 	 Incident Data Base. This data file contains records of cases received from news
    clips, medical examiner reports, consumer complaints, and reports from other
    sources. CPSC officials reviewed cases reported during the previous 10 years
    (1993 through 2002).

Regarding these databases, CPSC officials noted that statistics on police shootings
involving toy guns likely represent an undercounting of such incidents. The officials
explained that CPSC would not generally collect data on police shootings because the
agency’s focus is on consumer product issues rather than firearms.

We contacted CDC to inquire whether its National Violent Death Reporting System
(NVDRS) had relevant information. Officials at CDC’s National Center for Injury



Page 9                                                      GAO-03-1135R Use of Toy Guns
Enclosure I


Prevention and Control told us that the system was not yet operational but eventually
would collect national information on all violent deaths, including firearms-related
deaths and those involving toy guns.

The CDC officials referred us to Harvard University’s National Violent Injury
Statistics System (NVISS), which is a pilot program for CDC’s NVDRS and
encompasses 12 sites nationwide.6 The co-director of the pilot program told us that
the NVISS database contains information for 2 years (2000 and 2001) but does not
include a variable to facilitate an electronic search for injuries or deaths involving toy
gun incidents. Nevertheless, the co-director responded to our questions based on her
knowledge of the database and her review of data for 7 sites for 2001.7

Long-Term Impacts Attributed to Toy Gun Play by Children

As previously indicated, to determine the availability of studies examining the long-
term impacts that can be attributed to toy gun play by children, we conducted a
literature search and contacted various federal agencies, university researchers or
academicians, and a representative of the Toy Industry Association. We focused our
analysis on the following potentially relevant study that we identified:

• 	 Watson, Malcolm W., and Ying Peng. Brandeis University. “The Relation Between
    Toy Gun Play and Children’s Aggressive Behavior.” Early Education and
    Development (Volume 3, Number 4, Oct. 1992): 370-389.

Two of our social scientists examined the study to assess the adequacy of samples
and measures employed, the reasonableness and rigor of the statistical techniques
used to analyze them, and the validity of the results and conclusions.

Also, we contacted the senior scientific editor of a 2001 report—Youth Violence: A
Report of the Surgeon General—to discuss the extent, if any, that the study
addressed or considered the long-term impacts attributable to toy gun play by
children.8




6
  The 12 NVISS sites consist of 6 states (Connecticut, Kentucky, Maine, Maryland, Wisconsin, and Utah); 5 metropolitan areas or 

counties—Atlanta (Ga.), Detroit (Mich.), Miami-Dade County (Fla.), Allegany County (Pa.), and San Francisco (Cal.)—and 

1 pilot (conducted by the University of Pennsylvania) collectively covering three metropolitan areas—Bethlehem (Pa.),

Youngstown (Ohio), and Iowa City (Iowa).

7
  The co-director’s review covered the following 7 sites: Connecticut, Maine, Wisconsin, Utah, San Francisco, Miami-Dade

County, and Allegany County. 

8
  Youth Violence: A Report of the Surgeon General, U.S. Public Health Service (Washington, D.C.: Jan. 2001). The report is 

available on the Surgeon General’s Web site at http://surgeongeneral.gov/library/youthviolence/.




Page 10                                                                                  GAO-03-1135R Use of Toy Guns
Enclosure II

             June 1990 Report on Toy Guns, Crime, and Police Encounters

Public Law 100-615 (Nov. 5, 1988) required that toy guns have a “blaze orange plug
inserted in the barrel” to minimize the probability of such guns being mistaken for
real firearms. Also, the federal legislation required that the Department of Justice’s
Bureau of Justice Statistics (BJS) conduct a study of the criminal misuse of toy guns,
including studying police reports of such incidents. Effective June 1, 1989, BJS
awarded a cooperative agreement to the Police Executive Research Forum (PERF) to
conduct the study. 9 In June 1990, PERF reported on the results of its study—Toy
Guns: Involvement In Crime & Encounters With Police.

Regarding the overall significance of toy guns and crime, PERF reported that there is
no clear answer to the question, “How serious is the problem?” The report noted that:

•    “The response is a value judgment based upon one’s ideology and experiences.”

• 	 “In comparison to all crimes of violence and police-involved shootings throughout
    the United States, the proportion of cases involving imitation guns is small. The
    nagging element of the ‘toy gun’ problem is that many of the incidents seem
    particularly tragic—a child is involved, a mentally disturbed person does not
    recognize the gravity of his/her actions, or a person simply used poor judgment.”

In response to our inquiry, in June 2003, BJS officials told us that BJS had no current
plans to sponsor or undertake a follow-up study to update the information.

Scope and Methodology of PERF’s Study

Initially, to ascertain issues and trends, PERF conducted a Lexis/Nexis computer
search for news stories reporting any imitation gun incidents. PERF reported that its
research methods also included developing and sending a survey instrument to 699
agencies in the study population—that is, all municipal police and consolidated
police departments serving populations of 50,000 or more, all sheriff’s departments
with 100 or more sworn employees, and all primary state police agencies.10 According
to PERF, the usable response rate was 65.5 percent (458 responses). The PERF
report does not include information on whether the nonrespondents differed in
significant ways from the respondents. Without such information, it is not possible to
determine if the lack of response from 34 percent of the agencies distorted the
findings.



9
 PERF is a national membership organization of police executives from the largest city, county, and state law enforcement 

agencies. Incorporated in 1977, PERF's objectives are to improve policing and advance professionalism through research and 

involvement in public policy debate.

10
   Also, according to PERF, site visits were made to 27 agencies, which were selected based partly on news reports and/or self-

reports indicating experiences with imitation gun incidents. 




Page 11                                                                                   GAO-03-1135R Use of Toy Guns
Enclosure II


Questions in the survey instrument solicited information on (1) robberies and
assaults that involved the use of imitation guns and (2) experiences of officers using
deadly force and less than deadly force against individuals with imitation guns.11

Data Limitations Acknowledged by PERF

In its report, PERF acknowledged that quantitative data were difficult to obtain.
Specifically, PERF reported that:

“The questions for this survey were extraordinarily difficult for the law enforcement agencies to
answer simply because police departments typically do not maintain data stratified by the identifying
character of ‘toy gun’ (or similar notation). As a result, most agencies resorted to some form of manual
records check and/or solicitation of information from officers.”

“The data presented in this report were not easily generated by the responding departments. Many
police agencies conducted manual searches of their incident reports, others physically searched
property room records, while others went through the laborious process of surveying their officers and
then developing responses to our questions. Thus, while the data in this report may not be as robust as
we initially hoped, it represents the most comprehensive information available on the subject.”

Nonetheless, relying solely on officers’ memories in some instances may have
resulted in either an under- or over-reporting of incidents involving imitation guns.


Study Results Reported by PERF

As mentioned previously, questions in PERF’s survey instrument solicited
information on (1) robberies and assaults that involved use of imitation guns and
(2) experiences of officers using deadly force and less than deadly force against
individuals with imitation guns.

Robberies and Assaults Using Toy Guns

PERF’s report presented data for 4-2/3 years—1985 through 8 months of 1989—on the
numbers of robberies and assaults involving toy guns. As table 2 shows, the reported
number of robberies totaled 2,796 during this period.




11
  PERF reported results separately for three categories of imitation guns—toy guns (intended for playing), pneumatic guns (such
as BB and pellet guns), and replica guns (inoperable reproductions of actual weapons). The data presented in tables 2 and 3 in
this enclosure cover toy guns only.



Page 12                                                                                 GAO-03-1135R Use of Toy Guns
Enclosure II


Table 2: Number of Robberies and Assaults Committed Involving Toy Guns (Jan. 1, 1985 to Sep. 1, 1989)

                            Number of robberies and assaults committed involving toy guns
    Calendar year                             Robberiesa                        Assaultsb
    1985                                              453                              635
    1986                                              482                              615
    1987                                              665                              601
    1988                                              753                              686
    1989 (8 months)                                   443                              567
    Total                                           2,796                            3,104
Source: Police Executive Research Forum, Toy Guns: Involvement in Crime & Encounters with Police, June 1990, pp. 29 and
32.
a
Based on 148 police agencies reporting robberies known to have been committed with a toy gun.
b
Based on 121police agencies reporting assaults known to have been committed with a toy gun.

Based partly on these data, PERF reported that survey results “show that robberies
by imitation guns are occurring on a daily basis in the United States …”. Also, PERF
reported that:

“Because of poor record keeping on imitation gun robberies, the fact that the estimates of
investigators are experiential rather than empirical, and the inherent methodological differences
between the UCR [Uniform Crime Reporting Program] and this study, the authors feel that estimating
the number of imitation gun robberies from those reported in the UCR would have limited value.”

Also, table 2 shows that the reported number of assaults committed with toy guns
totaled 3,104 during the period. PERF reported that:

“While it is conceivable that a person could be physically assailed with an imitation gun, the more
likely crime is the ‘simple assault’ where a person is threatened and in fear of injury. … No
meaningful comparisons can be made between these findings and the Uniform Crime Report assault
data since the UCR statistics reflect only aggravated assaults.”

Use of Force Incidents Involving Toy Guns

As part of PERF’s survey, researchers also asked law enforcement agencies to report
the number of incidents where officers had warned, threatened, or actually used
force in a confrontation where an imitation gun had been mistaken for a real gun. As
table 3 shows, for the period January 1, 1985, to September 1, 1989:

• 	 The number of incidents totaled 385 where an officer warned or threatened the
    use of force based on the belief that a toy gun was real.

• 	 The number of incidents totaled 105 where an officer used actual force based on
    the belief that a toy gun was real.




Page 13                                                                             GAO-03-1135R Use of Toy Guns
Enclosure II


Table 3: Number of Incidents Where Police Officers Warned of Using or Actually Used Force Based on
the Belief That a Toy Gun Was Real (Jan. 1, 1985 to Sep. 1, 1989)


                                                                             Number of incidents where an officer
                             Number of incidents where an officer            used actual force (deadly or less than
                            warned or threatened the use of force             deadly) based on the belief that a toy
                                                                    a                                              b
Calendar year           based on the belief that a toy gun was real                                  gun was real
1985                                                             55                                                6
1986                                                             61                                                7
1987                                                             72                                               45
1988                                                            106                                               31
1989 (8 months)                                                  91                                               16
Total                                                           385                                             105
Source: Police Executive Research Forum, Toy Guns: Involvement in Crime & Encounters with Police, June 1990, pp. 34 and
35.
a
Based on 82 agencies reporting incidents known to have been committed with a toy gun.
b
Based on 31 agencies reporting incidents known to have been committed with a toy gun.


Regarding the 105 incidents involving actual use of force, PERF reported that “… it is
probable that the data are more accurate since internal investigations typically follow
the use of force.”




(440210) 





Page 14                                                                             GAO-03-1135R Use of Toy Guns