Q1a. Please more clearly distinguish
the contribution of the CSLLEA, particularly at its current periodicity.
Unlike
the Uniform Crime Reports (UCR) and Census of Governments (COG), the CSLLEA
provides an opportunity to collect additional data to supplement the basic
personnel counts. This includes a breakdown of agency functions, and how both
sworn and civilian personnel are allocated to functions such as patrol,
investigations, court security, process serving, and jail operations. As is planned for the 2008 data
collection, data that provide some context to the trends observed in personnel
counts can also be included.
Examples of this include questions related to recruitment and
retention.
Q1b. What are the major scope and
methodological differences between the law enforcement employment counts in
CSLLEA and the Census Bureau�s quinquennial Census of Governments?
The
annual counts of sworn personnel in the UCR and COG (as well as the Expenditure
and Employment (E & E) sample survey based on the COG), like those from the
CSLLEA, include only full-time employees with arrest powers. However, the CSLLEA
includes all officers with arrest powers regardless of function, while the UCR
data exclude officers not paid out of police funds. This generally excludes
employees working for jails or courts. A similar exclusion is found in the
sample data from the Census Bureau's Annual Survey of Public Employment
published by
Q1c. How similar are the estimates
over time?
Generally
the employment trends reported by the
CSLLEA, UCR, and the COG/E & E have been similar over time. However, the UCR
and COG/E & E totals have been
consistently lower than those generated by the CSLLEA. For example, in 2004, the
CSLLEA reported 1,076,897 full-time personnel employed by law enforcement
agencies, compared to 970,588 for the UCR, and 892,436 for the E and E. Counts
are also higher for full-time sworn personnel, with a CSLLEA total of 731,903
compared to 675,734 for the UCR and 688,909 for the E and E.
Another
reason the UCR counts are lower than those from the CSLLEA is that the UCR
excludes some agencies that do not have an Originating Agency Identification
(ORI) number assigned by the FBI. Some agencies without an ORI are still
included in the UCR employee counts (but not in the agency counts), because they
report their data to another agency which in turn reports it to the FBI.
Overall, the UCR data cover 95% of the
Q1d. How suitable would the Census
of Governments be to serve as a sampling frame for LEMAS?
A
major limitation of the COG for use as a sampling frame for LEMAS is that it is
a government-based survey whereas the CSLLEA and LEMAS are agency-based. The government offices surveyed in the
COG would not be an appropriate contact list for the broader data sought from
the State police, local police, sheriffs� offices, and other law enforcement
agencies responding to the CSLLEA.
A sampling procedure based on the
COG would also be limited due to the existence of multiple agencies within a
single government�s budget. For example a State government�s employment totals
for police protection might include the State Highway Patrol, State Park Police,
and numerous State university and college police. This complicates the development of an
agency sample. The converse is also true where a single agency serves across
jurisdictional boundaries such as a transit or airport authority police
department.
2. Please clarify who the
current CSLLEA collection agent is and whether it is a grant-funded collection,
as indicated in A14.
The
current cooperative agreement award for the 2008 CSLLEA is with the National
Opinion Research Center (NORC) at the