2012 LEMAS SUPPORTING STATEMENT - Part A v2

SUPPORTING STATEMENT

2012 Law Enforcement Management and Administrative Statistics (LEMAS) Survey

Overview

The Bureau of Justice Statistics (BJS) proposes to implement the 2012 Law Enforcement Management and Administrative Statistics (LEMAS) Survey.¹ The proposed survey is a substantively revised and enhanced effort designed to build upon and improve the contributions of the previous eight waves of the LEMAS surveys.^. Like previous LEMAS surveys, the proposed 2012 survey is a nationally-representative sample of approximately 3,500 state and local general purpose law enforcement agencies in the United States; unlike previous LEMAS surveys, the design of the proposed survey will introduce a new core-and-supplement model. This request is for data collection that will occur during the period October 1, 2012 through June 30, 2013.

This statistical series generates national estimates about the characteristics of the 12,500 state and local general purpose law enforcement agencies; the functions they perform; the resources available to them; the number, types, and working conditions of their employees; the automation of agency functions and their information systems; the extent to which weapons are authorized and used; the formal policies that guide and restrict the behavior of sworn personnel; and the organizational responses utilized by these agencies to address contemporary law enforcement challenges.

Experience with Surveying General Purpose Law Enforcement Agencies, 1987 - 2007

LEMAS is part of a program of law enforcement statistics that has traditionally emphasized surveys of organizations. The core of the program is the one-page Census of State and Local Law Enforcement Agencies (OMB 1121-0240) conducted every four years since 1992. This census documents the number of agencies and the numbers of sworn personnel in those agencies. The census also provides the basis for distinguishing among various types of agencies by asking the types of functions performed by the agencies (e.g., law enforcement, investigative, court security, jail management, and process serving). In addition to the general purpose agencies surveyed by LEMAS, the Census identifies a variety of special purpose agencies. The single largest type of special purpose law enforcement agencies are campus police which BJS surveyed in 1995 and 2005. The 2012 Survey of Campus Law Enforcement Agencies (OMB 1121-0334) is currently in the field².

BJS also collects data from law enforcement organizations and from medical examiners and coroners about deaths that occur during the process of arrest. Since 2003 BJS reports the number of homicides, suicides, accidental deaths and deaths due to intoxication or medical conditions by year, state, type of agency and type of death as part of the Deaths in Custody Reporting Program (OMB 1121-0249).

Since 1987, BJS has successfully implemented eight waves of LEMAS surveys. These surveys captured information about the changing aspects of law enforcement organization, resources, functions, personnel, salaries, training, collective bargaining, information systems, policies, and use of technology. Certain topics have been covered in each wave; other topics have been dropped due to lack of interest or the difficulty in obtaining reliable answers.

A recent review of BJS programs by the National Research Council (NRC) recognized the crucial place of the LEMAS surveys in the BJS statistical programs but criticized this program for its limited focus on administrative and managerial characteristics of law enforcement agencies (Groves and Cork, 2009). The NRC report recommended several changes to this program. First, the academy recommended that BJS law enforcement surveys should collect more information about law enforcement agency behavior and performance. Second, BJS should enhance the use of agency identifiers to encourage the linkage of agency-specific organizational characteristics with agency specific-crime statistics and with the demographic characteristics of the jurisdictions served by each agency. Third, noting the lengthy instrument and the irregular schedule of past LEMAS surveys, the NRC recommended that BJS adopt a “core and supplement” design for a regularly scheduled program of agency surveys. The NRC suggested the consistent use of a limited number of core items that would be integrated with thematic supplements which would vary from wave to wave. Each of these recommended changes have been incorporated into the proposed 2012 LEMAS Survey.

During a meeting of LEMAS stakeholders held in December 2011 (see Section A Part 8), attendees identified two technical aspects that plagued prior waves of LEMAS. In prior waves of LEMAS, many response options were a single checked box, which engendered confusion between missing data and a “no” response. Therefore, BJS included explicit “yes,” “no,” and “N/A” categories where appropriate to help eliminate such confusion in the 2012 LEMAS. The second concern was that many topics, especially policy related topics, were addressed with a single survey item; this raised questions about the reliability of such survey responses. For example, several policy questions on prior LEMAS surveys asked only whether an agency had a written policy on a particular topic (e.g., deadly force). Stakeholders suggested that asking for additional information about specific polices would likely improve item reliability for the upcoming 2012 LEMAS. For example, question G6 –departmental foot pursuit policy—was expanded from asking whether a department had a foot pursuit policy to asking if the policy included certain specific elements (e.g., containment tactics).

Design of the 2012 LEMAS Survey

BJS proposes to build upon and enhance the LEMAS statistical series. The 2012 instrument has been constructed with a core set of questions and several thematic supplements. The format of several survey items has been improved to address concerns about measurement issues. The sampling plan has been revised to enhance the efficiency and precision of national estimates. The agency and jurisdiction identifiers recommended by the NRC for linking LEMAS information to data from the Federal Bureau of Investigation’s (FBI) Uniform Crime Reporting Program (UCR) and census-based demographic data were added in the 2007 LEMAS survey data file (BJS, 2011) and will be retained in the 2012 LEMAS and future BJS surveys of law enforcement agencies.

Web-based data collection will be emphasized in the 2012 LEMAS to promote high response rates, speed data collection, and simplify data verification and report preparation. The revised survey instrument and the new sampling design combined with shortened periods for data collection and report production provide the basis for establishing a shorter cycle for future BJS surveys of general purpose law enforcement agencies. BJS has selected the Urban Institute to act as the data collection agent for this program. The Urban Institute will collect various paradata (e.g., respondent response mode, time required to answer each question, total time for survey completion, the time interval between respondent access to the survey and completion of the survey, etc.) that will allow BJS to better evaluate the success of the online data collection effort. This information will also enable BJS to develop strategies to encourage greater online data collection for future LEMAS surveys. Additionally, BJS plans to test differences in response rates for online versus paper based data collection modes on a related project (2012-R2-CX-K017), the 2012 Census of State and Local Law Enforcement Agencies (CSLLEA). Working with the CSLLEA contractor, BJS proposes to conduct a response mode experiment. If OMB approves the project as proposed, some respondents will initially be given access to the online instrument and would need to request a paper instrument if desired, while the other group of respondents will initially be mailed a paper survey with information detailing how they can directly access the web based survey rather than completing the paper survey. Results from each group will be compared to determine if there are differences in response rates for the two response modes. The findings from this experiment will further BJS’ effort to move future waves on LEMAS data collections online.

A. Justification

1. Necessity of Information Collection

Under Title 42, United States Code, Section 3732 (see Attachment 1), the Bureau of Justice Statistics (BJS) is directed to collect and analyze statistical information concerning the operation of the criminal justice system at the federal, state, and local levels. State and local general purpose law enforcement agencies are the primary point of entry into the criminal justice system. Law enforcement agencies play a crucial gate keeping function in receiving reports of offenses, investigating crimes and making arrests.

In the United States, local law enforcement agencies are numerous and diverse. In 2008, there were 15,564 local police agencies and sheriff offices. Most of these organizations had fewer than 10 sworn personnel but the largest 409 agencies employed half of all sworn personnel. The functions, policies, and practices of local law enforcement agencies are determined and implemented by local governments with limited state-level coordination and oversight. State law enforcement agencies are few in number, large in size and typically emphasize a limited range of law enforcement functions, such as traffic enforcement. Because of the diversity and number of independent state, local and county governments, there is no organizational basis for systematically collecting and regularly reporting changes in the characteristics of their law enforcement agencies or the personnel those agencies employ, except for BJS-sponsored surveys of law enforcement agencies.³

The behavior of law enforcement agencies and employees are regularly subjected to extensive scrutiny from within each agency, by external government agencies, by the public and by the media. Much of what is communicated to the public about these agencies stems from media reports of dramatic or unusual events in which an officer, suspect or innocent bystander is injured or killed or incidents in which law enforcement officials are (or appear to be) violating laws, departmental policies, or social expectations. Media reports of heroic actions by law enforcement personnel to protect human life and property can communicate a similarly skewed accounting of the nature of law enforcement in the United States. The salience of these unusual events can provide the public and public policymakers with an incomplete and inaccurate understanding of the nature of state and local law enforcement organizations and their personnel.

Another limited perspective on law enforcement comes from surveys conducted by university professors and law enforcement professional associations (e.g., Police Executive Research Forum, International Association of Chiefs of Police). These surveys are typically conducted among only a small number of large law enforcement agencies and, even among this limited sample, have survey response rates that rarely exceed fifty percent. The appearance of objectivity in academic or professional surveys is reduced when survey responses are used to support advocacy positions of the survey sponsor or when the details of the survey instrument, data collection, and data analysis are not publicly available for independent review.

The past LEMAS surveys have not, and the proposed 2012 LEMAS will not, collect information about the nature of criminal behavior or injuries to sworn personnel. These two information needs are addressed by the FBI. In its Uniform Crime Reporting Program and Supplemental Homicide Reporting Program (SHR), the FBI collects detailed information about the nature of crimes reported to State and local law enforcement agencies. In its Law Enforcement Officers Killed and Assaulted (LEOKA) and Supplemental Homicide Reporting (SHR) Programs, the FBI collects detailed information about the number of law enforcement officers killed or assaulted. The 2012 LEMAS will include FBI ORI codes that can be used to link LEMAS data on agency characteristics with FBI data on known offenses, arrest, as well as officers killed or assaulted.

2012 LEMAS Survey Items

The proposed 2012 LEMAS survey will collect detailed information on the number of full-time and part-time, sworn and nonsworn, paid and volunteer personnel in law enforcement agencies. For full-time sworn personnel, the survey will collect data on the race and sex of officers. In addition, LEMAS is the only source for information about the nature and variety of salary levels, employment benefits, pension programs, and education requirements for sworn personnel. The need for objective information about a wide range of agencies characteristics derived from a representative sample of law enforcement agencies is addressed by no other source but the BJS program of law enforcement agency surveys in general and the LEMAS Survey in particular.

The 2012 LEMAS focuses on a core set of questions about the characteristics of law enforcement agencies and their personnel and five sets of supplemental questions. Attachment 2 provides a list of each survey item by core and supplement domains and whether the specific item had been included in prior BJS law enforcement organizational surveys.

Core

The core items in the 2012 LEMAS capture basic descriptive information about the name, address and agency ORI code needed to link these responses to past and future law enforcement organizational surveys. The core items also include the number of full-time and part-time sworn and nonsworn personnel in each agency. For full-time sworn personnel, the 2012 LEMAS records the number of employees by race and sex. The 2012 LEMAS captures the extent to which sworn personnel are included in traditional retirement program, individual retirement programs or social security. Two other essential items in the core are the number of sworn and nonsworn personnel that were hired and the number that were separated from the agency in the past year. The core of the survey also includes the agency’s annual operating budget; the salary schedule for entry level, first line supervisors and chief executives; and whether agency has a collective bargaining agreement with sworn personnel. The last set of core survey items is the organizational response to a set of specific problems or challenges facing law enforcement, such as a specialized unit for bias or hate crimes or dedicated personnel trained to use special weapons and tactics. A version of the organizational response survey item has been asked in every LEMAS survey since 1987 as a measure of organizational specialization.

Supplement: Agency Resources

The items in the first supplement in the 2012 LEMAS Survey capture the extent to which agencies resources, employment standards, personnel benefits and base salaries have been impacted by changing policies and economic conditions. This includes items about furloughs and pay freezes since 2010 and the amount of overtime work authorized or available in the past year. In this component, the 2012 LEMAS captures the nature and extent of any hiring freezes and the changes in the educational requirements for newly hired sworn personnel.

Supplement: Community Policing

The community policing component of the 2012 LEMAS Survey includes 9 survey items that have been used in prior LEMAS surveys since 1999. These items capture the nature of an agency’s mission statement, use of problem solving techniques, engagement in community collaboration, officer patrol assignments, surveys of residents, and community policing training. The use of these items in four prior LEMAS surveys will permit the 2012 LEMAS to document the trends in these items over a 15 year period.

Supplement: Information Systems

Information systems are the lifeblood on modern law enforcement agencies. The 2012 LEMAS captures details about the technologies available to officers to receive information while they are in the field. These items are similar to information obtained in prior LEMAS surveys. In addition, the 2012 LEMAS requests never before obtained information about how agencies collect, maintain and analyze data about criminal incidents. These set of survey items will help BJS document the capabilities of state and local law enforcement agencies to analyze incident level crime information, an essential element in contemporary crime reduction strategies.

Supplement: Officer Safety

Recent increases in assaults on, and in deaths of, sworn law enforcement officers enhance the need for systematic data on the policies and practices of law enforcement agencies concerning officer safety. The survey items included in this supplement document departmental policies for purchasing and using body armor and for pursuits on foot and in motor vehicles.

Supplement: Uses of Force

Prior LEMAS surveys have documented the range of tactics and weapons that are authorized by law enforcement agencies. The 2012 LEMAS Survey extends this focus to include items about how agencies document uses of force, what types of force must be documented, and the total number of use of force incidents reported in the past year.

2. BJS Needs and Uses

The traditional model of the criminal justice system begins with victims of crime reporting to law enforcement agencies. Reports from these agencies are typically (but not exclusively) the basis for all future considerations of the criminal justice system. Because of this strategic role, the program of law enforcement agency surveys is a core data collection for BJS. The 2012 LEMAS Survey provides the only systematic and objective basis to produce national estimates of personnel, resources, functions, policies, and practices of the most common type of law enforcement agency.

BJS has and will continue to use the LEMAS surveys to produce information available from no other source. With LEMAS, BJS can produce systematic national estimates on full-time and part-time sworn and nonsworn law enforcement personnel (regardless of whether or not those agencies participate in the FBI’s Uniform Crime Reporting Program). The LEMAS surveys have provided, and will continue to provide, the only national measures of the demographic composition of sworn officers; the salaries, benefits, and working conditions for sworn law enforcement personnel; and the extent to which these agencies employ nonsworn personnel.

These data are needed by BJS (and by the nation) to understand the extent to which law enforcement personnel are representative of the communities they serve and to determine if the working conditions of these personnel are sufficient and appropriate for the responsibilities they face. In addition, this survey provides BJS with systematic knowledge about the resources, policies practices, and organizational responses used to meet the challenges faced by contemporary law enforcement agencies. Comparisons of the 2012 LEMAS with prior LEMAS surveys will also provide important information on how law enforcement agencies have been affected by the recession.

Without LEMAS, BJS will be unable to describe the number and types of officers in state and local law enforcement agencies and to report to the nation what law enforcement agencies do. For example, LEMAS provides concrete measures of the extent to which general purpose law enforcement agencies are using various components of the community policing model recommended by various agencies of the U. S. Department of Justice. LEMAS also provides BJS with new information about the current capacity of specific law enforcement agencies to conduct location-specific crime analyses and to link that capacity to an agency’s historic and contemporary rates of reported crime. In addition, the 2012 LEMAS is needed to provide BJS with nationally representative data on current policies to enhance officer safety, the number of vehicle pursuits, and the nature and amount of force used. These issues are at the center of public policy discussions and are included in the 2012 LEMAS to permit BJS to produce national estimates of the contemporary status of law enforcement agencies as well as reports on changes over time in these matters.

In thirty reports BJS staff has used LEMAS surveys to describe a large number of characteristics of different types of law enforcement agencies (e.g. large police departments, sheriff offices). These reports are often cited in textbooks, research articles and public discussions as the authoritative source on the characteristics of state and local law enforcement agencies. BJS staff have also used LEMAS data to produce reports on thematic issues such as use of force complaints (Hickman and Piquero, 2009), women in law enforcement (Langton, 2010), and the comparison of campus and city police operations (Bromley and Reaves, 1998). (See Attachment 3 for a complete listing of publications derived from LEMAS data.)

Uses of the LEMAS Surveys by Others

The information generated from the LEMAS surveys is widely used and cited by the law enforcement professional and research communities. As of January 2012, there have been over 100 published reports derived from the eight waves of LEMAS data. The private publications authored by independent researchers tend to use the LEMAS data⁴ in conjunction with other sources of information to address specific topics such as police arrest decisions, law enforcement uses of geographic information systems, and law enforcement responses to specific issues (e.g., hate crimes, gangs, intimate partner violence). The LEMAS data are also used by justice department officials (U.S. Department of Justice, 2011) and in widely read publications by professional law enforcement organizations (Melekian, 2012) as authoritative statistics on law enforcement trends.

A frequent user of the information obtained from the LEMAS surveys is the staff from state and local law enforcement agencies, often individuals from the same agencies that complete the LEMAS surveys. While some of these agencies are interested in summary statistics or national averages provided by BJS published reports, BJS frequently answers inquiries from law enforcement personnel about aspects of a select number of agencies that the inquiring agencies consider their peers. For instance, the Phoenix Police Department may want to compare itself with the Los Angeles Police Department or a small sheriff’s office in Alabama may want to know how many other sheriff offices in Alabama have a gang unit or a use of force policy. The existence of specialized units or the adoption of new technology are two aspects of the LEMAS surveys which have been a regular interest to law enforcement personnel, many of whom are considering creating new units or purchasing similar equipment for their agencies.

Two other frequent users of the LEMAS data are the media and the public. The BJS Law Enforcement Statistics Unit answers hundreds of calls every year from the public or from media reporters. These calls often concern details that can be answered only with information from BJS surveys of general purpose law enforcement agencies.

The revised design of the LEMAS Survey will enhance the use of these data by law enforcement professionals and researchers. First, by archiving data files with consistent agency identification numbers for each agency, analysts outside of BJS can more easily examine changes in particular agencies or groups of agencies over time. Second, the 2012 LEMAS results will be linked, through ORI codes, with data from the FBI’s UCR and LEOKA data at the agency level. Lastly, LEMAS incorporates data about agency performance as well as the characteristics of organizations and personnel. These aspects of the 2012 LEMAS should increase external interest in, and use of, the LEMAS data beyond that found in the previous LEMAS surveys.

Anticipated Products

BJS anticipates producing multiple reports from the 2012 LEMAS. The first report will emphasize the resources available to law enforcement agencies in 2012 and, where appropriate data exist, trends in those resources since 1987. This report will emphasize changes in the number of agencies, sworn and nonsworn personnel, agency budgets and equipment, salary schedules, and employment benefits. For many of these items, we will have eight waves of data for reporting changes over time. This report will also address the recent use of furloughs, layoffs, and changes in benefits, overtime work and officer authority to work overtime for other agencies.

The 2012 LEMAS supplemental items on agency resources, community policing, information systems, officer safety and police use of force will be the basis for additional reports. Each of these reports will be short, focused on a single theme, and will incorporate national estimates for 2012 along with analyses of trends in policies and practices. For example, since virtually all of the community policing items in the 2012 LEMAS were used in LEMAS surveys since 1999, the report on community policing will emphasize trends in community policing. About half of the questions concerning officer use of information systems were present in previous LEMAS surveys and that report will blend reporting about contemporary activities and historical trends. Most of the officer safety and police use of force questions are new to LEMAS and the trend analyses in these reports will be limited. [Attachment 3 provides a listing of survey items, their domains and the prior use of these items in LEMAS or other BJS surveys of law enforcement agencies.]

In addition to the planned analytical reports, BJS will produce an on-line data tool using all the data from the 2012 LEMAS so that law enforcement professionals, law enforcement researchers and the general public can have immediate access to information about individual law enforcement agencies as well as summary information by agency type, size, or location.

At the time of the initial publication from the 2012 LEMAS, BJS will release fully-documented data files for public use through the National Archive of Criminal Justice Data at the University of Michigan.

3. Use of Information Technology

The 2012 LEMAS instrument has been designed for online data collection that will export survey data and paradata in SPSS format. This software will allow the Urban Institute to send an email to respondents explaining the LEMAS program and containing a hyperlink to the questionnaire. Additionally, the software allows for real-time online tracking of respondents thereby allowing BJS to track the completion of each agency’s responses.

Some law enforcement organizations without access to the internet or lack the capabilities to respond to an electronic questionnaire will be sent a paper-based survey by fax or through the mail. These data will be entered into the automated data file as they are received noting the date and method of data submission.

The dataset, and supporting documentation, will be made available for download without charge at the National Archive of Criminal Justice Data at the Inter-University Consortium for Political and Social Research (ICPSR) and at Data.gov. Access to these data permits analysts to identify the specific responses of individual agencies and to conduct statistical analyses about general purpose law agencies. These data will have agency and jurisdiction specific identifiers that will permit the public use of these data in combination with other data files with similar agency or location identifiers.

The BJS-produced findings from the 2012 LEMAS Survey will be provided to the public in electronic format. These reports will be available on the BJS website as PDF files. BJS will also produce a web-based, data analysis tool for the 2012 LEMAS Survey to increase the ease with which the public can access information about specific agencies or types of agencies.

4. Efforts to Identify Duplication

Based on our knowledge of the Federal statistical system, in general, and law enforcement surveys in particular, BJS has determined that the 2012 LEMAS includes measures of the number of law enforcement personnel and their salaries that are also included in two ongoing surveys by other Federal agencies.⁵ The Federal Bureau of Investigation annually collects information from law enforcement agencies about the number and sex of sworn and nonsworn personnel as part of the “Number of Full-Time Law Enforcement Employees” (OMB No. 1110-0004). The Bureau of Labor Statistics’ (BLS) “Occupational Employment Survey” (OMB No. 1220-0042) samples employers yearly about the number and salaries of employees in three Protective Service Occupation subcategories: 1) police and sheriff patrol officers, 2) detectives and criminal investigators, and 3) first line supervisors of police and detectives.

BJS has identified four variables—the number of male sworn, male nonsworn, female sworn, and female nonsworn personnel—that are collected and reported by the FBI survey and by BJS in its Census of State and Local Law Enforcement Agencies (OMB NO. 1121-0240) and in the LEMAS surveys. In the five years (1992, 1996, 2000, 2004 and 2008) for which both the FBI survey and the BJS census were conducted, the FBI collected data from 3,600 to 5,200 fewer agencies (24.9%) and report about 100,000 fewer total personnel (10.0%). These differences are due in part to the different criteria for inclusion of agencies and personnel in these two surveys. The FBI survey is limited to personnel paid “with law enforcement funds” while the BJS surveys include all personnel regardless of what funds pay their salaries. In addition, BJS survey captures all agencies that employ the equivalent of at least one full-time sworn personnel; the FBI survey is limited to agencies that report to the FBI’s UCR program during that particular year (See Reaves, 2011).

BJS and the FBI collect and report personnel numbers at the agency level. They use slightly different definitions of “law enforcement” personnel and capture different proportions of the total population of law enforcement agencies and personnel. The few data items about personnel in the FBI survey are collected in conjunction with annual data collections of hundreds of items about reported offenses and about assault on law enforcement officers and are needed to report offense, arrest and assault rates per sworn personnel.

The items about personnel in the BJS law enforcement census and in the LEMAS surveys are used to produce national estimates of personnel and to provide the basis for computing the percentages of sworn personnel by race and ethnicity, by law enforcement function and by current and newly hired personnel. LEMAS also collects information about part-time, seasonal and volunteer employees of law enforcement agencies. The number of duplicate data collection items in the BJS and FBI data collection is small and the information collected is necessary to meet the internal needs of each survey.

Turning to the Occupational Employment Survey, both the BJS and BLS surveys report information about the number of law enforcement employees and their salaries. The BLS survey emphasizes comparisons of the number of positions and their compensation among many occupations types across different geographical areas of the country. The samples and employee definitions used in these two surveys vary widely depending upon the varied purposes of the surveys. In law enforcement surveys, the distinction between sworn and nonsworn is crucial but this distinction can only be assumed in the BLS occupational sub codes. Moreover, many law enforcement employees, such as forensic scientists or crime analysts, are unlikely to fit into any BLS occupational codes for protection service occupations. Lastly, the BLS survey is concerned with the range of salaries actually paid in SMSAs and LEMAS is concerned with the lowest and highest salaries possible for entry level personnel in particular agencies and groups of agencies.

As with the FBI survey, the number of duplicate items in the BJS and BLS surveys is small and the items are needed for the internal purposes of the survey. The BJS annual data are collected and reported at the agency level and at the national level separately for sheriff offices and police departments. The BLS data are collected at the employer level and three year averages are reported at the SMSA level and the national level with no distinction among Federal, state or local law enforcement agencies.

BJS has identified three Federally-sponsored surveys with varying samples and measures of employees that can be used to estimate the number of sworn law enforcement personnel in the United States. However, only the BJS surveys have as one of their primary purposes the production of national estimates of the number of law enforcement agencies and the number of sworn and nonsworn personnel. As part of the 2012 LEMAS program, BJS will conduct diagnostic tests of these three surveys and collaborate with the FBI and the BLS to better understand the characteristics of these three measures of law enforcement personnel.

Another way that BJS has worked to avoid duplication is by coordinating with the staff at the National Institute of Justice (NIJ) responsible for its programs on body armor and crime analysis to insure that the items used in this survey do not duplicate their current or planned efforts and that it will provide useful information for the future of their programs. In particular, the scope of the items about crime analysis were co-developed between BJS and NIJ to enhance the value of this survey and the proposed 2013 NIJ in-depth survey of crime analysis in large law enforcement agencies. In addition, the revised community policing supplement in the 2012 LEMAS stems from collaborations with the Community Oriented Policing Service (COPS) of the U.S. Department of Justice. This collaboration has assisted BJS, NIJ, and the COPS office in avoiding the collection of information already available or soon to be available from surveys or other data collections conducted by each agency.

5. Efforts to Minimize Burden

The proposed 2012 LEMAS instrument was designed to reduce the respondents’ burden in three ways. First, the instrument reduces the number of specific information items requested from 324 to 268, and those items are organized into a limited number of topical areas. Second, the 2012 instrument was designed for web-based data collection, with built-in assistance modules and edit checks. Third, based on the guidance BJS has received from the law enforcement professional and research communities, the 2012 LEMAS instrument has improved the wording of traditional and new survey items.

We expect that most respondents will make use of the online survey software to complete the survey. A number of web-based system functions will be in place to ease the burden of survey completion. The Urban Institute will utilize an intelligent log-in program for data collection, which will store agency information and responses, allowing for multi-session, non-sequential completion of the survey instrument. Since many agencies, particularly the larger ones, will need to seek out multiple information sources within their organizations to answer different sections, this will reduce the burden on them by facilitating data entry from different sources. It will also reduce the burden by allowing them to stop response entry pending confirmation of information from others in the agency. Help icons located next to each survey question will link respondents to item-specific information, additional guidance, and helpdesk contact information to facilitate requests for assistance.

The online system will also provide a glossary of terms for respondent reference. In addition, a Help Desk will be staffed during normal business hours (east coast time) and will be available to respondents through a toll free number. Respondents who lack the capabilities to access and utilize the web-based survey instrument will receive a paper-based survey by fax or mail accompanied with paper based definitions of terms and directions. BJS will also create a help desk that will provide assistance by phone and email to all paper-based respondents.

A data collection manager will oversee the help desk; when not available, calls will automatically be routed to another survey team member for immediate response. Voice mail will be available during off hours and a dedicated LEMAS help e-mail address will be provided with the introductory letter and survey packet. The office and cell-phone numbers, as well as e-mail address for the survey principal investigator, will also be provided to respondents to insure timely communications.

In addition, BJS has sought guidance from law enforcement professionals and research institutions. A principal objective of these consultations was the production of a clear, accurate survey in order to minimize respondent burden. Input from these consultations about the difficulty of certain questions has been incorporated into the current instrument to simplify questions, improve response code options and clarify technical language and jargon used in law enforcement agencies.

In March of 2012, BJS tested a draft instrument with 9 law enforcement agencies and the interviews conducted with these agencies resulted in changes in the number, type and wording of questions on the proposed survey instrument.

6. Consequences of Less Frequent Collection

Based in part on recommendations from the National Research Council (Groves and Cork, 2011) and the Director of the Office of Community Oriented Policing (Melekian, 2011), BJS has determined that it is necessary to improve the timeliness of its law enforcement data collections and to establish a more regular schedule of future surveys of general purpose law enforcement agencies. Following the completion of the 2012 LEMAS in 2013, BJS will evaluate alternative approaches for scheduling future waves of this program, from regular waves every two years to an ongoing annual data collection. BJS has funded a separate solicitation (2012-R2-CX-K011), the BJS Analytic Resource Center (BJS-ARC) project, which will help with this evaluation. The BJS-ARC is designed to provide scientific and technical support for statistical and methodological research, statistical analyses, documentation, and dissemination services in support of BJS’s Criminal Justice Statistics Program (CJSP). LEMAS is one of the initial data collections identified in the BJS-ARC project to undergo evaluation. One aspect of the BJS-ARC project will be to identify and evaluate potential LEMAS sampling waves (e.g. administering LEMAS every two years) and/or the utility of making LEMAS an ongoing data collection similar to the U.S. Census Bureau’s American Community Survey (ACS). The BJS-ARC project will also establish a working group of academic and policing professionals, which would meet on a regular basis, to evaluate the substantive content of LEMAS, including any proposed supplemental topics. Working with stakeholders will allow BJS to identify more quickly those emerging issues in the field of law enforcement that may be included in future versions of the LEMAS survey.

7. Special Circumstances

No special circumstances have been identified for this project.

8. Adherence to 5 CFR 1320.8(d) and Outside Consultations

BJS shared a copy of a draft LEMAS survey instrument with research scholars with a known interest in law enforcement issues at the 2011 Annual Meeting of the American Society of Criminology. In addition, the draft instrument was shared with five national law enforcement professional associations—the International Association of Chiefs of Police, the National Sheriffs Association, the Major City Chiefs Association, the Police Executive Research Forum and the Police Foundation--and with staff at the FBI, the National Institute of Justice and the COPS Office.

On questions about employee pensions and benefits, BJS consulted with Holly Deal, a researcher at the National Fraternal Order of Police, and with Keith Brainard, Research Director at the National Association of State Retirement Administrators. In addition, in December 2011, BJS hosted a stakeholders meeting for the upcoming LEMAS survey. Participants included representatives from various law enforcement agencies, research organizations, and universities. Representatives from the COPS Office and the International Association of Chiefs of Police were also in attendance. Participants discussed a variety of topics, including survey content, data availability and use by the field, methods to maximize response and ways to minimize respondent burden. Attendees at the working group meeting are listed on the following page.

9. Paying Respondents

Neither BJS nor the Urban Institute will provide any payment or gift of any type to respondents. Respondents will participate on a voluntary basis.

Attendees of the 2012 LEMAS Working Group
Stephen Mastrofski, Professor George Mason University 4400 University Drive Fairfax, VA 22030	John Kapinos, Strategic Planner Fairfax County Police Department 4100 Chain Bridge Road Fairfax, VA 22030
Gary Cordner, Professor Kutztown State University Old Main 367 15200 Kutztown Rd. Kutztown, PA 19530	Ed Maguire, Professor Justice, Law and Society American University 4400 Massachusetts Avenue NW Washington, DC 20016
William King, Associate Professor Sam Houston State University Box 2296 Huntsville, TX 77341	Cynthia Lum, Professor George Mason University 4400 University Drive Fairfax, VA 22030
Christine Eith, Assistant Professor Johns Hopkins University Police Executive Leadership Columbia Center 6740 Alexander Bell Drive Columbia, MD 21046	Howard Snyder, Deputy Director Bureau of Justice Statistics U.S. Department of Justice 810 Seventh Street NW Washington, DC 20531
Matthew Scheider, Assistant Director Office of Community Oriented Policing U.S. Department of Justice 145 North S Street NE Washington, DC 20530	Joel Garner, Chief, Bureau of Justice Statistics U.S. Department of Justice 810 Seventh Street NW Washington, DC 20531
John Markovic, Program Manager Office of Community Oriented Policing U.S. Department of Justice 145 North S Street NE Washington, DC 20530	Ron Malega, Statistician Bureau of Justice Statistics U.S. Department of Justice 810 Seventh Street NW Washington, DC 20531
John Firman, Director of Research International Association of Chiefs of Police 515 North Washington Street Alexandria, VA 22314	David Hayeslip, Principal Investigator Justice Policy Center The Urban Institute 2100 M Street NW Washington, DC 20037
Stephen Bamford, Captain, Investigative Services Manassas Police Department 9518 Fairview Avenue Manassas, VA 20110	Samantha Lowry, Project Director Justice Policy Center The Urban Institute 2100 M Street NW Washington, DC 20037

10. Assurance of Confidentiality

According to 42 U.S.C. 3735 Section 304, the information gathered in this data collection shall be used only for statistical or research purposes, and shall be gathered in a manner that precludes their use for law enforcement or any purpose relating to a particular individual other than statistical or research purposes. The data collected through the LEMAS Survey represent institutional characteristics of publicly-administered law enforcement agencies. The information about these organizations is in the public domain. The fact that participation in this survey is voluntary and that information about individual agency responses will be available to the public is included on the first page of the survey instrument. However, BJS will not release the names, phone numbers or email of the actual persons responsible for completing the 2012 LEMAS survey instruments.

11. Justification for Sensitive Questions

There are no questions of a sensitive nature in the proposed 2012 LEMAS Survey.

12. Estimate of Respondent Burden

BJS has estimated the respondent burden for the proposed 2012 LEMAS Survey at 10,744 hours. This estimate is based on four factors—the estimated burden for the 2007 LEMAS survey, the number and type of questions asked in the 2007 survey and in the proposed 2012 survey, the burden estimates generated from 9 tests of an earlier draft survey instrument, and changes made to that draft in the final instrument proposed here.

According to the supporting statement for the 2007 LEMAS survey a burden of 7,400 hours was

“based on experience garnered from previous administrations of the LEMAS survey, pre-testing conducted as part of those previous administrations, as well as consideration of the currently proposed instrument”.

The 2007 burden estimate was derived from an estimate of 3 hours for the completion of the 10 page long-form version of the 2007 LEMAS survey by 1,000 agencies and an estimate of 2 hours for the completion of the 8 page short-form version of the 2007 LEMAS survey by 2, 200 agencies (3 hours X 1,000 plus 2 hours X 2,200 = 7,400). The long form of the 2007 LEMAS survey included the collection of data on 324 variables. Seventy-six of those variables required the respondents to report of an amount (personnel, hours, dollars, etc.) and 227 required the respondent to check a single item. The short form version of the 2007 LEMAS survey included 213 variables, 73 required a count and 140 required a check.

In March 2012, a paper-based version of the draft 2012 LEMAS instrument was pilot tested with eight state and local police departments and one sheriff’s office. The agencies ranged in size from around 60 to over 1,300 sworn law enforcement officers. Respondents were asked to complete the survey, consider the clarity of survey questions and provide time burden estimates for each of 9 sections of the survey. Within a week of their completing these pilot surveys, respondents were interviewed by phone with specific questions about each section and the overall survey.

We were able to get useable estimates of burden from 8 of the 9 sites piloting this survey (see Table 1). The burden estimates vary greatly from agency to agency and from section to section. One agency estimated that their burden was more than 12 hours; another agency reported less than 1 hour. Section G of the survey instrument concerning vehicles and pursuit consumed almost 50% of the burden for one agency but less than six percent of the burden for 6 other agencies. In general, Section A through D, with most of the count questions, took longer than sections E through I.

On average, the eight respondents reported that it took 4.35 hours to complete that draft instrument. The three agencies with less than 100 sworn officers estimated an average burden of 3.00 hours; the five agencies with more than 100 sworn officers averaged 5.14 hours. In addition to their section-by-section estimate of the time needed to complete this survey, the nine agencies provided BJS with extensive feedback on the specific wording and organization of the survey instrument. The respondents in this 9 agency test reported what is well known in survey methodology, that items requiring a count are more burdensome than items that merely require a “Yes or “No” response or a check off of one among many options. In addition, they identified a few word selections (e.g., “furlough”, “hiring freeze”, “seasonal employee”) that unexpectedly warranted clarification or definitions. Respondents complained about the many racial categories and the complications of counting all motorized vehicles in an agency. One responding agency noted that we need to ask about “across the board” permanent reductions in the base salary for all employees, not just one time furloughs for a few days or weeks.

In response to these comments and to the previous burden estimates, BJS revised its draft instrument in several ways. First, we added a question about “across the board” salary decreases and we clarified language in questions or in response codes for a dozen survey items. Second, we determined that, while the 8 response codes for the race of officers did add burden, the value of the racial details warranted the burden. Third, we dropped 11 variables completely; these variables were questions added to enhance the information we had about other items. For instance, the tested instrument asked about percent contribution of employers and employee to pension programs. While this additional information would provide new details on the nature of employee compensation, we determined that their burden exceeded the value of the additional information sought. Eliminating these variables reduced burden but eliminated some potentially useful details about particular issues in policing. Fourth, we converted the response codes in 34 count variables into yes/no responses. For instance, we retained the question about motorized vehicles that had been asked in prior LEMAS surveys but requested detailed count information on only two types of vehicles and used less burdensome check boxes for the use of other types of vehicles.

The tested instrument had 279 variables, 119 of which required a count and 160 that required only a check. The revised instrument has 268 variables, of which 85 require respondents to estimate a count. The total number of variables included in the final instrument is reduced by 5% but the number of count variables is reduced by 29%. In addition, the tests were conducted with paper forms. BJS will utilize online survey software for most respondents and this is likely to reduce the burden on respondents.

Table 1: Estimated Time to Complete February 2012 Draft of the LEMAS Survey (in minutes)
		Agencies
		1	2	3	4	5	6	7	8	Total
Sections	A	60	120	20	15	2	30	15	45	307
	B	45	45	30	10	15	30	20	30	225
	C	45	120	45	30	5	30	15	20	310
	D	20	120	60	10	10	25	20	10	275
	E	20	60	40	10	5	15	10	15	175
	F	45	60	60	10	5	10	10	15	215
	G	30	60	15	120	2	20	5	10	262
	H	45	60	30	30	5	5	10	10	195
	I	10	60	20	10	2	1	5	15	123
Total Minutes		320	705	320	245	51	166	110	170	2,087
Hours		5.3	11.8	5.3	4.1	0.9	2.8	1.8	2.8	4.3
Number of Employees		310	201	61	456	18	79	105	1,331

The 2012 LEMAS Survey will be sent to approximately 3,500 state and local law enforcement agencies. Based on the 4.35 hour average completion time in the pilot tests, the amount of burden based on the piloted instrument is estimated at 15,225 hours. If the burden estimate takes into account the larger number of smaller agencies in this survey and the smaller burden reported by smaller agencies, the burden estimate for the piloted instrument is estimated at 12,640 (see Table 2).

The final instrument collects data on 11 fewer variables (5%) than the tested version and converts 56 count variables (29%) into less burdensome check off variables. In addition, the 2012 LEMAS will be implemented in a less burdensome web-based format. Based on these changes from the tested version and the version that will be implemented in the field, BJS anticipates an additional 15% reduction in burden. Based on these calculations, our final estimate of the total burden of the proposed 2012 LEMAS Survey is 10,744 hours.

13. Estimate of Respondent’s Cost Burden

BJS anticipates that one person per surveyed agency will complete the data collection instrument, with pay approximately equivalent to the GS-12 / 01 level ($71,901 per year). Based on the estimated time burden, the agency cost of employee time would be approximately $34.45 an hour. The total respondent employee time cost burden is estimated at $370,130.

There are no anticipated costs to respondents beyond the employee time expended during completion of the survey instrument and addressed in the above section. This expectation was further reinforced through the March 2012 pre-test survey; none of the nine responding agencies reported additional costs incurred by survey participation.

14. Costs to Federal Government

The total expected cost to the Federal Government for this data collection is $768,953 all to be borne by the BJS. This work consists of planning, developing the questionnaire, preparation of materials, collecting the data, evaluating the results, and generating the PDF and web based query reports. A BJS GS-Level 15 statistician will be responsible for overseeing the Urban Institute’s work on this project. The budget for this project is shown in Table 4.

15. Reason for Change in Burden

The total estimated respondent time burden has increased by 3,162 hours from the estimated burden for the 2007 LEMAS survey. This change in burden is due, in part, to the elimination of the short form version of the LEMAS program survey. While small agencies will receive the same form as large agencies in the 2012 LEMAS, the pretest of the survey instrument showed that their burden will be lower than that for larger agencies (see Table 3). This variation in burden appears to stem from the fact that the response of smaller agencies to some items—such as the number and characteristics of nonsworn personnel—may be a simple “not applicable” in the 2012 LEMAS. The change in burden is also the result of a slightly larger sample size. The burden estimate for the 2007 LEMAS survey was based on a sample size of 3,200 agencies. If the 2012 LEMAS burden estimate were based on 3,200 agencies, the burden would be 9,979 hours for an increase of 2,579 hours.

Table 2: Estimated Burden Hours for the 2007 LEMAS,

2012 LEMAS Tested Instrument and

2012 LEMAS Reduced Instrument

		Estimated	Total
	Sample	Burden	Burden
	Size	(in hours)	Hours
	LEMAS 2007
Long Form	1,000	3.00	3,000
Short Form	2,200	2.00	4,400
Total Sample	3,200	2.3	7,400
	2012 LEMAS Tested Instrument
Total Sample	3,500	4.35	15,225
Large Agencies	1,000	5.14	5,140
Small Agencies	2,500	3.00	7,500
Weighted Sample	3,500	5.05	12,640
	2012 LEMAS Reduced Instrument
Large Agencies	1,000	4.37	4,369
Small Agencies	2,500	2.55	6,375
Weighted Sample	3,500	3.07	10,744

Table 3: Burden Estimate for 2012 LEMAS

Using Sample Size Used in 2007 LEMAS Estimate

	Sample Size	Estimated Burden (in hours)	Total Burden (in hours)
Large Agencies	1,000	4.37	4,369
Small Agencies	2,200	2.55	5,610
Weighted Sample	3,200	3.12	9,979

Another consideration is that the methodology for estimating the 2007 instrument’s burden was different than the methodology used to estimate the burden for the 2012 LEMAS instrument. While the 2012 estimate is derived from a more concrete process, the sample used was small and unrepresentative. The 2007 burden estimate was based on a more generalized experience of fielding a similar survey over 7 previous waves but was not directly linked to a test of the 2007 instrument.

Lastly, the increased burden of the 2012 LEMAS stems in part because BJS chose to go beyond measuring the characteristics of agencies and their employees to attempt to capture details about the behavior of law enforcement officers, as recommended by the National Research Council. While many of the 2012 LEMAS items are derived from the 2007 LEMAS instrument, the shift to measuring and reporting performance are the substantive changes which could be responsible for most of the increases in burden on respondents. The design of the 2012 LEMAS limits the burden of these performance measures. For instance, we ask only about the total number of use of force incidents, not the numbers for various types of force. Similarly, we considered but decided not to ask about the total number of personnel working as crime analysts, the characteristics of those personnel or the number of calls, offenses, or arrests.

16. Project Schedule

The data collection for 2012 LEMAS Survey is scheduled to begin in October 2012. The data collection period is 9 months. BJS plans to issue the first report from this project by September 1, 2013. BJS has determined that the shortened schedule for data collection is feasible because of paradata from the 2007 LEMAS data collection shows an 80% response rate within 6 months. In addition, BJS has provided substantial resources for the 2012 data collection and emphasized web-based data collection.

The design of the 2012 LEMAS program calls for the initiation of data analyses including the assessment of nonresponse biases when the response rate hits 50%. While this program anticipates a final response rate in excess of 90%, BJS is prepared to conduct and report national estimates based on response rates greater than 80%, if that is necessary to meet the deadline for the release of published findings by the end of September 2013.

Table 4: Estimated Costs for the 2012 LEMAS Survey
Bureau of Justice Statistics
Staff salaries
2012 Fiscal Year
GS-12 Statistician (20%)		$14,380
GS-15 Senior Statistician (15%)		$17,827
2013 Fiscal Year
Gs-12 Statistician (20%)		$14,380
GS-15 Senior Statistician (20%)		$17,827
GS-13 Editor (10%)		$8,550
Other Editorial Staff		$5,000
Senior BJS Management		$3,000
Subtotal salaries		$80,964
Fringe benefits (28% of salaries)		$22,670
Subtotal: Salary & fringe		$103,634
Other administrative costs of salary & fringe (15%)		$15,545
Subtotal: BJS costs		$119,179
Data Collection Agent (Urban Institute)
Personnel		$217,850
Fringe Benefits		$93.054
Travel		0
Consultant		$12,963
Contracts		$94,049
Other		$65,757
Total Indirect		$259,262
Subtotal Data Collection Agent		$649,974
Total estimated costs		$768,953

17. Display of Expiration Date

The expiration date will be shown on the survey form.

18. Exception to the Certificate Statement

BJS is not requesting an exception to the certification of this information collection.

1. Contacts for Statistical Aspects and Data Collection

1. BJS contacts include

• Howard Snyder

202-616-8305

[email protected]

• Ron Malega

202-353-0487

[email protected]

2. Persons consulted on statistical methodology:

• Rob Santos, The Urban Institute

3. Persons consulted on data collection and analysis:

• David Hayeslip, The Urban Institute

• Allen Beck, Bureau of Justice Statistics

202-616-3277

[email protected]

References Cited

Bureau of Justice Statistics. 2011. Law Enforcement Management and Administrative Statistics (LEMAS), 2007 [Computer file]. ICPSR31161-v1. Ann Arbor, MI: Inter-university Consortium for Political and Social Research [producer and distributor].

Bromley, Max and Brian Reaves. 1998. Comparing campus and city police operational practices. Journal of Security Administration 21 (2):41-54.

Ellis, Yukiko. 1997. Documentation of the Sample Selection for the 1997 Sample Survey of Law Enforcement Agencies. U.S. Bureau of the Census. Unpublished memorandum.

Groves, Robert and Daniel Cork. 2009. Ensuring the Quality, Credibility, and Relevance of U.S. Justice Statistics. Washington, D.C. National Academies Press.

Hickman, Matthew and Alex Piquero. 2009. Organizational, administrative, and environmental correlates of complaints about police use of force: Does minority representation matter? Crime and Delinquency 15 (1):3-27.

Kish, Leslie. 1965. Survey Sampling. New York. John Wiley and Sons.

Langton, Lynn. 2010. Women in Law Enforcement, 1987 – 2008. Washington, D.C.: Bureau of Justice Statistics.

Melekian, Bernard. 2012. Policing in the New Economy: A New Report on the Emerging Trends from the Office of Community Oriented Policing Services. Police Chief. 79 (January 2012): 16–19.

Reaves, Brian 2011. Census of State and Local Law Enforcement Agencies, 2008. Washington, D.C., Bureau of Justice Statistics.

Reaves, Brian, 2012. Federal Law Enforcement Officers, 2008. Washington, D.C., Bureau of Justice Statistics.

U.S. Department of Justice, 2011 The Impact of the Economic Downturn on American Police Agencies. Washington, D.C.; Community Oriented Police Service.

Attachments:

1. Title 42

2. Listing of Survey Items Identifying Core, Supplement and Trend Variables

3. Bibliography of Publications from LEMAS Surveys

4. 2012 Law Enforcement Management and Administrative Statistics Survey

5. Draft example of web version of LEMAS survey

6. Intro Letter from the BJS Director.

7. Draft Scripts for Respondent Contacts

8. 60-day ICR notice

9. 30-day ICR notice

1 In the 60-day and 30-day notices this work was called the Survey of General Purpose Law Enforcement Agencies (or SGPLEA). After comments from the field, the decision was made to return to the survey’s original title LEMAS.

2 The campus survey includes police departments run by private colleges and universities as well as departments operated by state and local government agencies.

3 The FBI collects some data on police personnel in its Uniform Crime Reporting Program, but these data differ from those collected by the LEMAS surveys in a number of ways that will be discussed later in this paper.

4 Beginning with the first LEMAS survey, BJS has provided free access to public-use data files at National Archive of Criminal Justice Data at the University of Michigan.

5 BJS’ Census of State and Local Law Enforcement Agencies, from which the LEMAS sample is drawn, also includes measures of the number of law enforcement personnel.

17