Methodology

Download Printable Document

Law enforcement agencies that participate in the Uniform Crime Reporting (UCR) Program forward crime data through their state UCR Programs in 46 states and the District of Columbia. Local agencies in those states that do not have a state Program submit crime statistics directly to the FBI, which provides continuous guidance and support to those participating agencies. The state UCR Programs function as liaisons between local agencies and the FBI. Many states have mandatory reporting requirements, and many state UCR Programs collect data beyond those typically called for by the national UCR Program to address crime problems specific to their particular jurisdictions. These state UCR Programs, in most cases, also provide direct and frequent service to their participating law enforcement agencies, make information readily available for statewide use, and help to streamline the national Program's (FBI's) operations.

Criteria for state UCR Programs

The criteria established for state Programs ensure consistency and comparability in the data submitted to the national Program, as well as regular and timely reporting. These criteria are:

The state Program must conform to the national UCR Program standards, definitions, and information required.
The state criminal justice agency must have a proven, effective, statewide Program and have instituted acceptable quality control procedures.
The state crime reporting must cover a percentage of the population at least equal to that covered by the national UCR Program through direct reporting.
The state Program must have adequate field staff assigned to conduct audits and to assist contributing agencies in record-keeping practices and crime-reporting procedures.
The state Program must furnish the FBI with all of the detailed data regularly collected by the FBI from individual agencies that report to the state Program in the form of duplicate returns, computer printouts, and/or appropriate electronic media.
The state Program must have the proven capability (tested over a period of time) to supply all the statistical data required in time to meet publication deadlines of the national UCR Program.

Data completeness and quality

In order to fulfill its responsibilities in connection with the UCR Program, the FBI continues to edit and review individual agency reports for both completeness and quality. Members of the national Program's staff contact the state UCR Program in connection with crime-reporting matters and, as necessary when approved by the state, individual contributors. Upon request, staff members conduct training programs within the state on law enforcement record-keeping and crime-reporting procedures. Following audit standards established by the federal government, the FBI conducts an audit of each state's UCR data collection procedures once every 3 years. Should circumstances develop whereby the state Program does not comply with the aforementioned requirements, the national Program may institute a direct collection of Uniform Crime Reports from law enforcement agencies within the state.

Reporting procedures

Law enforcement agencies tabulate the number of Part I (murder and nonnegligent manslaughter, forcible rape, robbery, aggravated assault, burglary, larceny-theft, motor vehicle theft, and arson) offenses brought to their attention based on records of all reports of crime received from victims, officers who discover infractions, or other sources, and submit them each month to the FBI either directly or through their state UCR Programs.

Unfounded offenses, clearances, and value of property—When, through investigation, an agency determines that complaints of crimes are unfounded or false, the agency eliminates that offense from its crime tally through an entry on the monthly report. The report also provides the total number of actual Part I offenses, the number of offenses cleared, and the number of clearances that involve only offenders under the age of 18. (Law enforcement can clear crimes in one of two ways: by the arrest of at least one person who is charged and turned over to the court for prosecution or by exceptional means—when some element beyond law enforcement's control precludes the arrest of a known offender.) Law enforcement agencies also submit monthly to the FBI the value of property stolen and recovered in connection with the offenses and detailed information pertaining to criminal homicide and arson.

Persons arrested—In addition to reporting Part I offenses, law enforcement agencies provide monthly to the UCR Program data on persons arrested for all other crimes except traffic violations. These arrest data include the age, sex, and race of arrestees for both Part I and Part II offenses. Part II offenses encompass all crimes, except traffic violations, that are not classified as Part I offenses.

Officers killed or assaulted—Law enforcement agencies also report monthly to the UCR Program information regarding law enforcement officers killed or assaulted, and yearly, the number of full-time sworn and civilian law enforcement personnel employed as of October 31.

Hate crimes—At the end of each quarter, law enforcement agencies report summarized data on hate crimes, i.e., specific offenses that were motivated by an offender's bias against the perceived race, religion, ethnic origin, sexual orientation, or physical or mental disability of the victim. Those agencies participating in the UCR Program's National Incident-Based Reporting System (NIBRS) submit hate crime data monthly.

Editing procedures

The UCR Program thoroughly examines each report it receives for arithmetical accuracy and for deviations in crime data from month to month and from present to past years that may indicate errors. The UCR staff members compare an agency's monthly reports with the agency's previous submissions and with those from similar agencies to identify any unusual fluctuations in an agency's crime count. Large variations in crime levels may indicate modified records procedures, incomplete reporting, or changes in the jurisdiction's geopolitical structure.

Evaluation of trends—Data reliability is a high priority of the FBI, which brings any deviations or arithmetical adjustments to the attention of state UCR Programs or the submitting agencies. Typically, FBI staff members study the monthly reports to evaluate periodic trends prepared for individual reporting units. Any significant increase or decrease becomes the subject of a special inquiry. Changes in crime reporting procedures or annexations that affect an agency's jurisdiction can influence the level of reported crime. When this occurs, the FBI excludes the figures for specific crime categories or totals, if necessary, from the trend tabulations.

Training for contributors—In addition to the evaluation of trends, the FBI provides training seminars and instructional materials on crime reporting procedures to assist contributors in complying with UCR standards. Throughout the country, the national Program maintains liaison with state Programs and law enforcement personnel and holds training sessions to explain the purpose of the Program, the rules of uniform classification and scoring, and the methods of assembling the information for reporting. When an individual agency has specific problems in compiling its crime statistics and its remedial efforts are unsuccessful, personnel from the FBI's Criminal Justice Information Services Division may visit the contributor to aid in resolving the difficulties.

UCR Handbook—The national UCR Program publishes a Uniform Crime Reporting Handbook (revised 2004), which details procedures for classifying and scoring offenses and serves as the contributing agencies' basic resource for preparing reports. The national staff also produces letters to UCR contributors and State Program Bulletins as needed. These provide policy updates and new information, as well as clarification of reporting issues.

The final responsibility for data submissions rests with the individual contributing law enforcement agency. Although the FBI makes every effort through its editing procedures, training practices, and correspondence to ensure the validity of the data it receives, the accuracy of the statistics depends primarily on the adherence of each contributor to the established standards of reporting. Deviations from these established standards that cannot be resolved by the national UCR Program may be brought to the attention of the Criminal Justice Information Systems Committees of the International Association of Chiefs of Police and the National Sheriffs' Association.

Population estimation

The FBI calculated 2006 state growth rates using revised 2005 state/national population estimates and 2006 provisional state/national population estimates provided by the U.S. U.S. Census Bureau. The FBI then estimated population figures for city and county jurisdictions by applying the 2006 state growth rate to the updated 2005 U.S. Census Bureau data.

NIBRS conversion

Thirty-one state Programs are certified to provide their UCR data in the expanded NIBRS format. For presentation in this book, the NIBRS data were converted to the historical Summary UCR data. The UCR Program staff constructed the NIBRS database to allow for such conversion so that UCR's long-running time series could continue.

Crime trends

By showing fluctuations from year to year, trend statistics offer the data user an added perspective from which to study crime. Percent change tabulations in this publication are computed only for reporting agencies that provided comparable data for the periods under consideration. The FBI excludes from the trend calculations all figures except those received for common months from common agencies. Also excluded are unusual fluctuations of data that the FBI determines are the result of such variables as improved records procedures, annexations, etc.

Caution to users

Data users should exercise care in making any direct comparison between data in this publication and those in prior issues of Crime in the United States. Because of differing levels of participation from year to year and reporting problems that require the FBI to estimate crime counts for certain contributors, the data are not comparable from year to year. In addition, this publication may contain updates to data provided in prior years' publications. Therefore, for example, the 2005 data in last year's publication may not match the 2005 data in this publication.

2006 Arrest data considerations

Because of changes in state or local agency reporting practices (updates to the NIBRS), figures are not comparable to previous years' data for Colorado (Denver).
Limited arrest data were received from Illinois (Chicago and Rockford only). However, the Illinois State UCR Program's guidelines for reporting forcible rape arrest counts do not comply with the national UCR Program's guidelines, i.e., Illinois data include arrests made for forcible rapes of male victims. Therefore, Illinois figures for forcible rape include only the forcible rape offenses from Rockford, submitted via NIBRS, that had female victims.
Because the Minnesota State UCR Program's guidelines for reporting forcible rape arrest counts do not comply with the national UCR Program's guidelines, i.e., Minnesota data include arrests made for forcible rapes of male victims, Minnesota figures for forcible rape, the violent crime total, and the total of all classes are not published.
For 2006, only arrest totals (with no age or gender breakdowns) are available for Florida.
No 2006 arrest data were received from the District of Columbia's Metropolitan Police Department; the only agency (Metro Transit Police) in the District of Columbia for which 12 months of arrest data were received has no attributable population.
No 2006 arrest data were received from Montana. However, arrest totals for this state were estimated by the national UCR Program and were included in Table 29 "Estimated Number of Arrests, United States, 2006."
No 2006 arrest data were received from the New York City Police Department. However, arrest totals for this area were estimated by the national UCR Program and were included in Table 29 "Estimated Number of Arrests, United States, 2006."

Offense estimation

Tables 1 through 5 and Table 7 of this publication contain statistics for the entire United States. Because not all law enforcement agencies provide data for complete reporting periods, the FBI includes estimated crime numbers in these presentations. The FBI estimates offenses that occur within each of three areas: Metropolitan Statistical Areas (MSAs), cities outside MSAs, and nonmetropolitan counties. The national Program computes estimates by using the known crime figures of similar areas within a state and assigning the same proportion of crime volumes to nonreporting agencies or agencies with missing data. The estimation process considers the following: population size of agency; type of jurisdiction, e.g., police department versus sheriff's office; and geographic location.

Estimation of state-level data

Various circumstances require the national Program to estimate offense totals for certain states. For example, some states do not provide forcible rape figures in accordance with UCR guidelines, or reporting problems at the state level have, at times, resulted in no usable data. In addition, the conversion of NIBRS data to Summary data has contributed to the need for unique estimation procedures. A summary of state-specific and offense-specific estimation procedures follows.

Year	State(s)	Reason for Estimation	Estimation Method
1986	Illinois	The state UCR Program was unable to provide forcible rape figures in accordance with UCR guidelines.	The forcible rape totals were estimated using national rates per 100,000 inhabitants within the eight population groups and assigning the forcible rape volumes proportionally to the state.
1987	Illinois	The state UCR Program was unable to provide forcible rape figures in accordance with UCR guidelines.	The forcible rape totals were estimated using national rates per 100,000 inhabitants within the eight population groups and assigning the forcible rape volumes proportionally to the state.
1988	Illinois	The state UCR Program was unable to provide forcible rape figures in accordance with UCR guidelines.	The forcible rape totals were estimated using national rates per 100,000 inhabitants within the eight population groups and assigning the forcible rape volumes proportionally to the state.
	Florida, Kentucky	Reporting problems at the state level resulted in no usable data.	State totals were estimated by updating previous valid annual totals for individual jurisdictions, subdivided by population group. Percent changes for each offense within each population group of the geographic divisions in which the states reside were applied to the previous valid annual totals. The state totals were compiled from the sums of the population group estimates.
1989	Illinois	The state UCR Program was unable to provide forcible rape figures in accordance with UCR guidelines.	The forcible rape totals were estimated using national rates per 100,000 inhabitants within the eight population groups and assigning the forcible rape volumes proportionally to the state.
1990	Illinois	The state UCR Program was unable to provide forcible rape figures in accordance with UCR guidelines.	The forcible rape totals were estimated using national rates per 100,000 inhabitants within the eight population groups and assigning the forcible rape volumes proportionally to the state.
1991	Illinois	The state UCR Program was unable to provide forcible rape figures in accordance with UCR guidelines.	The forcible rape totals were estimated using national rates per 100,000 inhabitants within the eight population groups and assigning the forcible rape volumes proportionally to the state.
	Iowa	NIBRS conversion efforts resulted in estimation for Iowa.	State totals were estimated by updating previous valid annual totals for individual jurisdictions, subdivided by population group. Percent changes for each offense within each population group of the West North Central Division were applied to the previous valid annual totals. The state totals were compiled from the sums of the population group estimates.
1992	Illinois	The state UCR Program was unable to provide forcible rape figures in accordance with UCR guidelines.	The forcible rape totals were estimated using national rates per 100,000 inhabitants within the eight population groups and assigning the forcible rape volumes proportionally to the state.
1993	Illinois	NIBRS conversion efforts resulted in estimation for Illinois.	Since valid annual totals were available for approximately 60 Illinois agencies, those counts were maintained. The counts for the remaining jurisdictions were replaced with the most recent valid annual totals or were generated using standard estimation procedures. The results of all sources were then combined to arrive at the 1993 state total for Illinois.
		The state UCR Program was unable to provide forcible rape figures in accordance with UCR guidelines.	The forcible rape totals were estimated using national rates per 100,000 inhabitants within the eight population groups and assigning the forcible rape volumes proportionally to the state.
	Kansas	NIBRS conversion efforts resulted in estimation for Kansas.	State totals were estimated by updating previous valid annual totals for individual jurisdictions, subdivided by population group. Percent changes for each offense within each population group of the West North Central Division were applied to the previous valid annual totals. The state totals were compiled from the sums of the population group estimates.
	Michigan, Minnesota	The state UCR Programs were unable to provide forcible rape figures in accordance with UCR guidelines.	The forcible rape totals were estimated using national rates per 100,000 inhabitants within the eight population groups and assigning the forcible rape volumes proportionally to each state.
1994	Illinois	NIBRS conversion efforts resulted in estimation for Illinois.	Illinois totals were generated using only the valid crime rates for the East North Central Division. Within each population group, the state's offense totals were estimated based on the rate per 100,000 inhabitants within the remainder of the division.
		The state UCR Program was unable to provide forcible rape figures in accordance with UCR guidelines.	The forcible rape totals were estimated using national rates per 100,000 inhabitants within the eight population groups and assigning the forcible rape volumes proportionally to the state.
	Kansas	NIBRS conversion efforts resulted in estimation for Kansas.	State totals were generated using only the valid crime rates for the West North Central Division. Within each population group, the state's offense totals were estimated based on the rate per 100,000 inhabitants within the remainder of the division.
	Montana	The state UCR Program was unable to provide complete 1994 offense figures in accordance with UCR guidelines.	State totals were estimated by updating previous valid annual totals for individual jurisdictions, subdivided by population group. Percent changes for each offense within each population group of the Mountain Division were applied to the previous valid annual totals. The state totals were compiled from the sums of the population group estimates.
1995	Kansas	The state UCR Program was unable to provide complete offense figures in accordance with UCR guidelines.	The state UCR Program was able to provide valid 1994 state totals which were then updated using 1995 crime trends for the West North Central Division.
	Illinois	The state UCR Program was unable to provide complete offense figures in accordance with UCR guidelines.	Valid Crime Index (Part I) counts were available for most of the largest cities (100,000 and over in population). For other agencies, the only available counts generated by the Illinois State Program were state totals based upon an incident-level system without indication of multiple offenses recorded within single incidents. Therefore, the UCR Hierarchy Rule could not be applied in order to convert the state's data to Summary data. (The Hierarchy Rule requires that only the most serious offense in a multiple-offense criminal incident is counted.) To arrive at a comparable state estimate to be included in national compilations, the Illinois State Program's state totals (which were inflated because of the nonapplication of the Hierarchy Rule) were reduced by the proportion of multiple offenses reported within single incidents in the NIBRS database. Valid totals for the large cities were excluded from the reduction process.
	Montana	The state UCR Program was unable to provide complete offense figures in accordance with UCR guidelines.	State estimates were computed by updating the previous valid annual totals using the 1994 versus 1995 percent changes for the Mountain Division.
1996	Florida	The state UCR Program was unable to provide complete offense figures in accordance with UCR guidelines.	The state UCR Program was able to provide an aggregated state total; data received from 94 individual Florida agencies are shown in the 1996 jurisdictional figures presented in Tables 8 through 11.
	Illinois	The state UCR Program was unable to provide complete offense figures in accordance with UCR guidelines.	Valid Crime Index (Part I) counts were available for most of the largest cities (100,000 and over in population). For other agencies, the only available counts generated by the Illinois State Program were state totals based upon an incident-level system without indication of multiple offenses recorded within single incidents. Therefore, the UCR Hierarchy Rule could not be applied in order to convert the state's data to Summary format. (The Hierarchy Rule requires that only the most serious offense in a multiple-offense criminal incident is counted.) To arrive at a comparable state estimate to be included in national compilations, the Illinois State Program's state totals (which were inflated because of the nonapplication of the Hierarchy Rule) were reduced by the proportion of multiple offenses reported within single incidents in the NIBRS database. Valid totals for the large cities were excluded from the reduction process.
	Kansas	The state UCR Program was unable to provide complete offense figures in accordance with UCR guidelines.	The Kansas state estimate was extrapolated from 1996 January-June state totals provided by the Kansas State UCR Program.
	Kentucky, Montana	The state UCR Programs were unable to provide complete offense figures in accordance with UCR guidelines.	The 1995 and 1996 percent changes within each geographic division were applied to valid 1995 state totals to generate 1996 state totals.
1997	Illinois	The state UCR Program was unable to provide complete offense figures in accordance with UCR guidelines.	Valid Crime Index (Part I) counts were available for most of the largest cities (100,000 and over in population). For other agencies, the only available counts generated by the Illinois State Program were state totals based upon an incident-level system without indication of multiple offenses recorded within single incidents. Therefore, the UCR Hierarchy Rule could not be applied in order to convert the state's data to Summary format. (The Hierarchy Rule requires that only the most serious offense in a multiple-offense criminal incident is counted.) To arrive at a comparable state estimate to be included in national compilations, the Illinois State Program's state totals (which were inflated because of the nonapplication of the Hierarchy Rule) were reduced by the proportion of multiple offenses reported within single incidents in the NIBRS database. Valid totals for the large cities were excluded from the reduction process.
	Kansas	The state UCR Program was unable to provide complete offense figures in accordance with UCR guidelines.	The Kansas state estimate was extrapolated from 1996 January-June state totals provided by the Kansas State UCR Program.
	Kentucky, Montana, New Hampshire, Vermont	The state UCR Programs were unable to provide complete offense figures in accordance with UCR guidelines.	The 1996 and 1997 percent changes registered for each geographic division in which the states of Kentucky, Montana, New Hampshire, and Vermont are categorized were applied to valid 1996 state totals to effect 1997 state totals.
1998	Delaware	The state UCR Program was unable to provide forcible rape figures in accordance with national UCR guidelines.	The 1998 forcible rape total for Delaware was estimated by reducing the number of reported offenses by the proportion of male forcible rape victims statewide.
	Illinois	The state UCR Program was unable to provide complete offense figures in accordance with UCR guidelines.	Valid Crime Index (Part I) counts were available for most of the largest cities (100,000 and over in population). For other agencies, the only available counts generated by the Illinois State Program were state totals based upon an incident-level system without indication of multiple offenses recorded within single incidents. Therefore, the UCR Hierarchy Rule could not be applied in order to convert the state's data to Summary format. (The Hierarchy Rule requires that only the most serious offense in a multiple-offense criminal incident is counted.) To arrive at a comparable state estimate to be included in national compilations, the Illinois State Program's state totals (which were inflated because of the nonapplication of the Hierarchy Rule) were reduced by the proportion of multiple offenses reported within single incidents in the NIBRS database. Valid totals for the large cities were excluded from the reduction process.
	Kansas	The state UCR Program was unable to provide complete offense figures in accordance with UCR guidelines.	To arrive at 1998 estimates, 1997 state totals supplied by the Kansas State UCR Program were updated using 1998 crime trends for the West North Central Division.
	Kentucky, Montana, New Hampshire, Wisconsin	The state UCR Programs were unable to provide complete offense figures in accordance with UCR guidelines.	State totals were estimated by using 1997 figures for the nonreporting areas and applying 1997 versus 1998 percentage changes for the division in which each state is located. The estimates for the nonreporting areas were then increased by any actual 1998 crime counts received.
1999	Illinois	The state UCR Program was unable to provide complete offense figures in accordance with UCR guidelines.	Valid Crime Index (Part I) counts were available for most of the largest cities (100,000 and over in population). For other agencies, the only available counts generated by the Illinois State Program were state totals based upon an incident-level system without indication of multiple offenses recorded within single incidents. Therefore, the UCR Hierarchy Rule could not be applied in order to convert the state's data to Summary format. (The Hierarchy Rule requires that only the most serious offense in a multiple-offense criminal incident is counted.) To arrive at a comparable state estimate to be included in national compilations, the Illinois State Program's state totals (which were inflated because of the nonapplication of the Hierarchy Rule) were reduced by the proportion of multiple offenses reported within single incidents in the NIBRS database. Valid totals for the large cities were excluded from the reduction process.
	Kansas, Kentucky, Montana	The state UCR Programs were unable to provide complete offense figures in accordance with UCR guidelines.	To arrive at 1999 estimates for Kansas, Kentucky, and Montana, 1998 state totals supplied by each state's UCR Program were updated using 1999 crime trends for the divisions in which each state is located.
	Maine	The state UCR Program was unable to provide complete offense figures in accordance with UCR guidelines.	The Maine Department of Public Safety forwarded monthly January through October crime counts for each law enforcement contributor; since 12 months of data were not received, the national Program estimated for the missing data following standard estimation procedures to arrive at a 1999 state total.
	New Hampshire	The state UCR Program was unable to provide complete 1999 offense figures in accordance with UCR guidelines.	The state total for New Hampshire was estimated by using the 1998 figures for the 1999 nonreporting areas and applying the 2-year percent change for the New England Division.
2000	Illinois	The state UCR Programs were unable to provide complete offense figures or forcible rape figures in accordance with UCR guidelines.	Valid Crime Index (Part I) counts were available for most of the largest cities (100,000 and over in population). For other agencies, the only available counts generated by the Illinois State Program were state totals based upon an incident-level system without indication of multiple offenses recorded within single incidents. Therefore, the UCR Hierarchy Rule could not be applied in order to convert the state's data to Summary format. (The Hierarchy Rule requires that only the most serious offense in a multiple-offense criminal incident is counted.) To arrive at a comparable state estimate to be included in national compilations, the Illinois State Program's state totals (which were inflated because of the nonapplication of the Hierarchy Rule) were reduced by the proportion of multiple offenses reported within single incidents in the NIBRS database. Valid totals for the large cities were excluded from the reduction process.
	Kansas	The state UCR Program was unable to provide complete offense figures in accordance with UCR guidelines.	To arrive at 2000 estimates for Kansas, 1999 state estimates were updated using 2000 crime trends for the West North Central Division.
	Kentucky, Montana	The state UCR Programs were unable to provide complete offense figures in accordance with UCR guidelines.	To arrive at 2000 estimates for Kentucky and Montana, 1999 state totals supplied by each state's UCR Program were updated using 2000 crime trends for the divisions in which each state is located.
2001	Illinois	The state UCR Program submitted complete data for only seven agencies within the state. Additionally, the state UCR Program was unable to provide forcible rape figures in accordance with UCR guidelines.	Valid Crime Index (Part I) counts were available for most of the largest cities (100,000 and over in population). For other agencies, the only available counts were generated without application of the UCR Hierarchy Rule. (The Hierarchy Rule requires that only the most serious offense in a multiple-offense criminal incident is counted.) To arrive at a comparable state estimate to be included in national compilations, the total supplied by the Illinois State Program (which was inflated because of the nonapplication of the Hierarchy Rule) was reduced by the proportion of multiple offenses reported within single incidents in the available NIBRS data. Valid totals for the large cities were excluded from the reduction process.
	Kentucky	The state UCR Program was unable to provide complete offense figures in accordance with UCR guidelines.	To arrive at the 2001 estimate for Kentucky, the 2000 state estimates were updated using 2001 crime trends reported for the East South Central Division.
2002	Illinois	The state UCR Program was unable to provide complete offense figures in accordance with UCR guidelines.	Valid Crime Index (Part I) counts were only available for most of the largest cities (100,000 and over in population). For other agencies, the only available counts generated by the Illinois State Program were state totals based upon an incident-level system without indication of multiple offenses recorded within single incidents. Therefore, the UCR Hierarchy Rule could not be applied in order to convert the state's data to Summary format. (The Hierarchy Rule requires that only the most serious offense in a multiple-offense criminal incident is counted.) To arrive at a comparable state estimate to be included in national compilations, the Illinois State Program's state totals (which were inflated because of the nonapplication of the Hierarchy Rule) were reduced by the proportion of multiple offenses reported within single incidents in the NIBRS database. Valid totals for the large cities were excluded from the reduction process.
	Kentucky	The state UCR Program was unable to provide complete offense figures in accordance with UCR guidelines.	To obtain the 2002 state crime count, the FBI contacted the state UCR Program, and the state agency was able to provide their latest state total, 2000. Therefore, the 2001 state estimate was updated for inclusion in the 2002 edition of Crime in the United States by using the 2001 crime trends for the division in which the state is located. To derive the 2002 state estimate, the 2002 crime trends for the division were applied to the adjusted 2001 state estimate.
2003	Illinois	The state UCR Program was unable to provide complete offense figures in accordance with UCR guidelines.	Valid Part I counts (formerly considered the Crime Index) were available only for most of the largest cities (100,000 and over in population). For other agencies, the only available counts generated by the Illinois State Program were state totals based upon an incident-level system without indication of multiple offenses recorded within single incidents. Therefore, the UCR Hierarchy Rule could not be applied in order to convert the state's data to Summary format. (The Hierarchy Rule requires that only the most serious offense in a multiple-offense criminal incident is counted.) To arrive at a comparable state estimate to be included in national compilations, the Illinois State Program's state totals (which were inflated because of the nonapplication of the Hierarchy Rule) were reduced by the proportion of multiple offenses reported within single incidents in the NIBRS database. Valid totals for the large cities were excluded from the reduction process.
	Kentucky	The state UCR Program was unable to provide complete offense figures in accordance with UCR guidelines.	To obtain the 2003 estimate, the 2003 crime trend for the East South Central Division was applied to an adjusted 2002 state estimate. The 2002 state count was reestimated by applying the 2002 crime trend for the East South Central Division using a more current figure, 2001 state totals, provided by the state UCR Program. The adjusted 2002 estimate differs from the figure published in the 2002 edition of Crime in the United States, which was originally estimated using 2002 state totals.
2004	Illinois	The state UCR Program was unable to provide complete offense figures in accordance with UCR guidelines.	Valid Part I counts were available only for agencies in the cities 100,000 and over in population. For other agencies, the only available counts generated by the Illinois State Program were totals based upon an incident-level system without indication of multiple offenses recorded within single incidents. Therefore, the UCR Hierarchy Rule could not be applied in order to convert the state's data to Summary format. (The Hierarchy Rule requires that only the most serious offense in a multiple-offense criminal incident is counted.) To arrive at a comparable state estimate to be included in national compilations, the Illinois State Program's totals (which were inflated because of the nonapplication of the Hierarchy Rule) were reduced by the proportion of multiple offenses reported within single incidents in the NIBRS database. Valid totals for the large cities were excluded from the reduction process.
2005	Illinois	The state UCR Program was unable to provide complete offense figures in accordance with UCR guidelines.	Valid Part I counts were available only for agencies in the cities 100,000 and over in population. For other agencies, the only available counts generated by the Illinois State Program were totals based upon an incident-level system without indication of multiple offenses recorded within single incidents. Therefore, the UCR Hierarchy Rule could not be applied in order to convert the state's data to Summary format. (The Hierarchy Rule requires that only the most serious offense in a multiple-offense criminal incident is counted.) To arrive at a comparable state estimate to be included in national compilations, the Illinois State Program's totals (which were inflated because of the nonapplication of the Hierarchy Rule) were reduced by the proportion of multiple offenses reported within single incidents in the NIBRS database. Valid totals for the large cities were excluded from the reduction process.
2006	Illinois	The state UCR Program was unable to provide complete offense figures in accordance with UCR guidelines.	Valid Part I counts (except for forcible rape [see below]) were available only for agencies in the cities 100,000 and over in population. For other agencies, the only available counts generated by the Illinois State Program were totals based upon an incident-level system without indication of multiple offenses recorded within single incidents. Therefore, the UCR Hierarchy Rule could not be applied in order to convert the state's data to Summary format. (The Hierarchy Rule requires that only the most serious offense in a multiple-offense criminal incident is counted.) To arrive at a comparable state estimate to be included in national compilations, the Illinois State Program's totals (which were inflated because of the nonapplication of the Hierarchy Rule) were reduced by the proportion of multiple offenses reported within single incidents in the NIBRS database. Valid totals for the large cities were excluded from the reduction process.
		The state UCR Program was unable to provide forcible rape figures in accordance with UCR guidelines.	Forcible rape figures for Rockford include only the forcible rape offenses with female victims that were extracted from the agency's NIBRS data. The rest of the state's forcible rape totals were estimated using national rates per 100,000 inhabitants within the eight population groups and assigning the forcible rape volumes proportionally to the state.
2006	Minnesota	The state UCR Program was unable to provide forcible rape figures in accordance with UCR guidelines.	The forcible rape totals were estimated using national rates per 100,000 inhabitants within the eight population groups and assigning the forcible rape volumes proportionally to the state.

Table Methodology

The tables in this report are based upon varying levels of data submissions. For example, some participating agencies may submit data for some but not all months of the reporting year. Using well-established procedures, the FBI estimates for missing offense data for agencies with partial reports and nonreporting agencies and then aggregates these estimates with data reported to determine the number of offenses for each state and the total U.S. population. Tables 17 and 29 present these approximations. In addition, various circumstances require the FBI to estimate offense totals from time to time for some states. (An explanation of the estimation procedures applied to particular states during specific reporting years is provided in the Offense Estimation section.)

To be included in Tables 8-11 and 21-22, which provide statistics for specific jurisdictions and states, agencies must submit 12 months of complete data prior to the FBI's established deadlines. To be included in Table 20, agencies must submit Supplementary Homicide Reports (SHRs). Tables 12-19 and 23-69 provide the number of reporting agencies (data source) and the total population covered by their collective jurisdictions.

The tabular presentation that follows briefly describes the data sources and the methods used to construct Tables 1-69.

(1) Table	(2) Database	(3) Table Construction	(4) General Comments
1- 1A	All law enforcement agencies participating in the UCR Program (including those submitting less than 12 months of data). Crime statistics for the Nation include all reported data and estimated offense totals (except arson) for agencies submitting less than 12 months of offense reports for each year.	The 2006 statistics are consistent with those published in Tables 2, 4, and 7. Prior to publication of this book, the FBI reestimated the state offense totals published in the previous edition of Crime in the United States to reflect data received after its publication. Because of this, the national totals for 2005 may have been adjusted. Population statistics represent the U.S. Census Bureau's July 1 provisional estimations for each year except 1990 and 2000, which are the decennial census years. (See the Population estimation.)	Represents an estimation of reported crime for the Nation from 1987 to 2006. The UCR Program does not have sufficient data to estimate arson offenses.
2	All law enforcement agencies participating in the UCR Program (including those submitting less than 12 months of data). Crime statistics for the Nation and for community types include all reported data and estimated offense totals (except arson) for agencies submitting less than 12 months of offense reports for each year.	The FBI estimates statistics for community types by aggregating the individual state statistics as shown in Table 5. Population estimates for 2006 are based on the percent change in state population from the U.S. Census Bureau's 2005 revised estimates and 2006 provisional estimates. (See the Population estimation.)	Represents an estimation of reported crime in 2006 for the: Nation MSAs Cities outside metropolitan areas Nonmetropolitan counties The UCR Program does not have sufficient data to estimate arson offenses.
3	All law enforcement agencies in the UCR Program (including those submitting less than 12 months of data). Crime statistics include all reported data and estimated offense totals (except arson) for agencies submitting less than 12 months of offense reports for each year.	The FBI computes regional offense distributions using the volume estimates as shown in Table 4. It bases population distributions on the U.S. Census Bureau's provisional estimates for 2006.	Represents the 2006 geographical distribution of estimated offenses and population. The UCR Program does not have sufficient data to estimate arson offenses.
4	All law enforcement agencies in the UCR Program (including those submitting less than 12 months of data). Crime statistics include all reported data and estimated offense totals (except arson) for agencies submitting less than 12 months of offense reports for each year.	The FBI derives state totals by estimating for nonreporting and partially reporting agencies within each state. Using the state's individual agency estimates, the Program aggregates the estimates with reported data to arrive at a state total. State totals for the prior year have been reestimated to reflect data received after the publication of the prior edition of Crime in the United States. Population statistics represent the U.S. Census Bureau's 2005 revised estimates and 2006 provisional estimates.	Represents an estimation of reported crime for the: Nation Regions Divisions States The UCR Program does not have sufficient data to estimate arson offenses. The Offense Estimation section supplies an explanation of the estimation procedures used for Illinois and Minnesota. Any comparisons of crime among different locales should take into consideration relevant factors in addition to the areas' crime statistics. Variables Affecting Crime provides more details concerning the proper use of UCR statistics.
5	All law enforcement agencies in the UCR Program (including those submitting less than 12 months of data). Crime statistics include all reported data and estimated offense totals (except arson) for agencies submitting less than 12 months of offense reports for each year.	Population estimates for 2006 are based on the percent change in state population from the U.S. Census Bureau's 2005 revised estimates and 2006 provisional estimates. (See the Population estimation.) Statistics under the heading Area Actually Reporting represent reported offense totals for agencies submitting 12 months of offense reports and estimated totals for agencies submitting less than 12 but more than 2 months of offense data. The statistics under the heading Estimated Totals represent the above plus estimated offense totals for agencies submitting 2 months or less of offense reports.	Represents an estimation of reported crime for states. The UCR Program does not have sufficient data to estimate arson offenses. The Offense Estimation section supplies an explanation of the estimation procedures used for Illinois and Minnesota. Any comparisons of crime among different locales should take into consideration relevant factors in addition to the areas' crime statistics. Variables Affecting Crime provides more details concerning the proper use of UCR statistics.
6	All law enforcement agencies in the UCR Program (including those submitting less than 12 months of data). Crime statistics include all reported data and estimated offense totals (except arson) for agencies submitting less than 12 months of offense reports for each year.	Table 6 includes all currently designated Metropolitan Statistical Areas (MSAs) in which at least 75 percent of the agencies within the MSA reported to the UCR Program and for which the principal city/cities submitted 12 months of complete data for 2006. Population estimates for 2006 are based on the percent change in state population from the U.S. Census Bureau's 2005 revised estimates and 2006 provisional estimates. (See the Population estimation.) The statistics under the heading Area Actually Reporting represent offense totals for agencies submitting 12 months of data and estimated totals for agencies submitting less than 12 but more than 2 months of data. The statistics under the heading Estimated Total represent the above plus estimated totals for agencies submitting 2 months or less of data.	Represents an estimation of reported crime for MSAs. The UCR Program does not have sufficient data to estimate arson offenses. Any comparisons of crime among different locales should take into consideration relevant factors in addition to the areas' crime statistics. Variables Affecting Crime provides more details concerning the proper use of UCR statistics.
7	All law enforcement agencies in the UCR Program (including those submitting less than 12 months of data). Crime statistics include all reported data and estimated offense totals (except arson) for agencies submitting less than 12 months of offense reports for each year.	The FBI estimates the breakdowns for robbery, burglary, and larceny-theft by first calculating the proportion of the total offense represented by each of the breakdowns as presented in Table 23 and applying those percentages to the estimated offense total as presented in Table 1.	Represents an estimation of reported crime for the Nation from 2002 to 2006. The data source from which the FBI derives Table 7 does not include aggravated assault or arson.
8	All town and city law enforcement agencies submitting 12 months of complete offense data for 2006.	The FBI estimated the 2006 population for each jurisdiction based on the percent change in the state population using the U.S. Census Bureau's 2005 revised estimates and 2006 provisional estimates. (See the Population estimation.) The FBI does not publish arson data unless it receives data from either the agency or the state for all 12 months of the calendar year.	Data users should take into consideration relevant factors in addition to the areas' crime statistics when making any valid comparisons of crime among different locales. Variables Affecting Crime provides more details concerning the proper use of UCR statistics.
9	All university/college law enforcement agencies submitting 12 months of complete offense data for 2006.	The 2005 student enrollment figures provided by the U.S. Department of Education are the most recent available. They include full and part-time students. The FBI does not publish arson data unless it receives data from either the agency or the state for all 12 months of the calendar year.	Represents reported crime from those individual college/university law enforcement agencies (listed alphabetically by state) contributing data to the UCR Program. Data users should take into consideration the size of enrollment, the number of on-campus residents, and other demographic factors when making any valid comparisons of crimes among colleges/universities. Variables Affecting Crime provides more details concerning the proper use of UCR statistics.
10	All county law enforcement agencies submitting 12 months of complete offense data for 2006.	The Metropolitan Counties classification encompasses jurisdictions covered by noncity law enforcement agencies located within currently designated MSAs. The Nonmetropolitan Counties classification encompasses jurisdictions covered by noncity agencies located outside currently designated MSAs. (See the Area Definitions.) The FBI does not publish arson data unless it receives data from either the agency or the state for all 12 months of the calendar year.	Represents reported crime from individual law enforcement agencies in metropolitan and nonmetropolitan counties (i.e., the individual sheriff's office and/or county police department). These figures do not represent the county totals because they exclude city crime counts. The state of Illinois did not contribute data for any county law enforcement agency. Data users should take into consideration relevant factors in addition to the areas' crime statistics when making any valid comparisons of crime among different locales. Variables Affecting Crime provides more details concerning the proper use of UCR statistics.
11	All state, federal, and territorial law enforcement agencies submitting 12 months of complete offense data for 2006.	State, federal, territorial, and other agencies are those agencies, regardless of jurisdiction, that are managed by their respective state, county, federal, or territorial governments or, in some cases, independently of those entities. The FBI does not publish arson data unless it receives data from either the agency or the state for all 12 months of the calendar year.	Represents reported crime from individual state or territorial law enforcement agencies (i.e., state police, highway patrol and/or other law enforcement agencies managed by the state or territory) and any federally managed law enforcement agency participating in the UCR Program. Data users should take into consideration relevant factors in addition to the areas' crime statistics when making any valid comparisons of crime among different locales. Variables Affecting Crime provides more details concerning the proper use of UCR statistics.
12-15	All law enforcement agencies submitting at least 6 common months of complete offense reports for 2005 and 2006.	The tables present 2-year comparisons based on 2005 and 2006 reported crime. In calculating trends, the UCR Program includes only common reported months for individual agencies. The FBI estimated the 2006 population for each jurisdiction based on the percent change in the state's population using the U.S. Census Bureau's 2005 revised estimates and 2006 provisional estimates. (See the Population estimation.) The Suburban and Nonsuburban Cities classification encompasses all cities other than principal cities served by municipal law enforcement agencies within MSAs.
16-19	All law enforcement agencies submitting 12 months of complete data (except arson) for 2006.	The FBI derived the offense rates by first dividing the total aggregated offense counts by the aggregated populations covered by contributing agencies and then multiplying the resulting figure by 100,000. The FBI estimated the 2006 population based on the percent change in state population from the U.S. Census Bureau's 2005 revised estimates and 2006 provisional estimates. (See the Population estimation.) UCR population breakdowns are furnished in the Area Definitions. The Suburban and Nonsuburban Cities classifications in Table 17 encompass all cities other than principal cities served by municipal law enforcement agencies in MSAs.	The data collection methodologies for the offense of forcible rape used by the state UCR Programs administered by the Illinois State Police and the Minnesota Department of Public Safety do not comply with UCR data collection guidelines. The FBI does not use the Illinois or Minnesota rape data in computing rates for Tables 16-18. The FBI computes the forcible rape rates in this table by estimating data for the agencies in Illinois and Minnesota that supply 12 months of complete data and applying the national rates of each represented population group. The UCR Program does not have sufficient data to estimate arson offenses. The totals provided in Table 19 reflect only those offenses for which law enforcement agencies provided additional information to the UCR Program; therefore, the totals will not match those shown in other rate tables.
20	All law enforcement agencies submitting Supplementary Homicide Report (SHR) data for 2006.	The weapon totals are aggregated from all murders for which the FBI received an SHR for calendar year 2006.	The SHR is the monthly law enforcement report to the UCR Program concerning homicides. It details victim and offender characteristics, circumstances, weapons used, etc. The SHR data submitted by Florida and Washington, D.C. did not meet UCR guidelines and were not included in this table.
21, 22	All law enforcement agencies submitting 12 months of complete offense weapon breakdown data for 2006.	The weapon totals are aggregated from all robberies (Table 21) and aggravated assaults (Table 22) for which the FBI received weapon breakdowns. Jurisdictional population statistics represent 2006 UCR estimates.	The FBI did not receive weapon breakdowns for offense data from Illinois.
23, 24	All law enforcement agencies submitting at least 6 months of complete property/classification data for 2006.	The FBI computes offense total and value lost total for all Part I offenses other than aggravated assault and arson. The percent distribution statistics are based on the offense total for each Part I offense. The FBI derives trends by comparing statistics from agencies with at least 6 common months of complete data reports for 2005 and 2006. (Offenses in Uniform Crime Reporting defines the UCR Program's Part I offenses.)	The offense of aggravated assault is not included in these tables. For UCR Program purposes, the taking of money or property in connection with an assault is reported as robbery. The data source from which the FBI derives Table 23 does not include arson.
25-28	All law enforcement agencies submitting at least 6 months of complete offense reports for 2006.	The FBI bases percent cleared statistics on aggregated offense and clearance totals. Population estimates for 2006 are based on the percent change in state population from the U.S. Census Bureau's 2005 revised estimates and 2006 provisional estimates. (See the Population estimation.) UCR population breakdowns are furnished in the Area Definitions.
29	All law enforcement agencies in the UCR Program (including those submitting less than 12 months of complete arrest data for 2006).	The arrest totals presented are national estimates based on the arrest statistics of all law enforcement agencies in the UCR Program (including those submitting less than 12 months of data). The estimated total number of arrests is the sum of known arrest figures plus estimated arrest volumes for each of the 28 offenses, not including suspicion. The arrest total for each of the individual offenses is the sum of the estimated volumes within each of the eight population groups. (See the Area Definitions.) The FBI calculated each group's estimate by dividing the reported volume figures (as shown in Table 31) by the contributing agencies' jurisdictional populations. The resulting figure was then multiplied by the total population for each population group as estimated by the Program. (See the Population estimation.)
30, 31	All law enforcement agencies submitting 12 months of complete arrest data for 2006.	The FBI derived the arrest rates by first dividing the total aggregated arrests by the aggregated populations covered by contributing agencies and then multiplying the resulting figure by 100,000. The population estimates for 2006 are based on the percent change in state population from the U.S. Census Bureau's 2005 revised estimates and 2006 provisional estimates. (See the Population estimation.) The Area Definitions contains the UCR population and geographical configuration.
32, 33	All law enforcement agencies submitting 12 months of complete arrest data for both 1997 and 2006.	The arrest trends are the percent differences between 1997 and 2006 arrest volumes aggregated from all agencies that submitted 12 months of arrest data for both years. The population estimates for 2006 are based on the percent change in state population from the U.S. Census Bureau's 2005 revised estimates and 2006 provisional estimates. Population estimates for 1997 are based on the percent change in state population from the U.S. Census Bureau's 1996 revised estimates and 1997 provisional estimates. (See the Population section.)
34, 35	All law enforcement agencies submitting 12 months of complete arrest data for both 2002 and 2006.	The arrest trends are the percent differences between 2002 and 2006 arrest volumes aggregated from all agencies that submitted 12 months of arrest data for both years. The population estimates for 2006 are based on the percent change in state population from the U.S. Census Bureau's 2005 revised estimates and 2006 provisional estimates. Population estimates for 2002 are based on the percent change in state population from the U.S. Census Bureau's 2001 revised estimates and 2002 provisional estimates. (See the Population estimation.)
36, 37	All law enforcement agencies submitting 12 months of complete arrest data for both 2005 and 2006.	The arrest trends are 2-year comparisons between 2005 and 2006 arrest volumes aggregated from agencies that submitted 12 months of arrest data in both years. The population estimates for 2006 are based on the percent change in state population from the U.S. Census Bureau's 2005 revised estimates and 2006 provisional estimates. Population estimates for 2005 are based on the percent change in state population from the U.S. Census Bureau's 2004 revised estimates and 2005 provisional estimates. (See the Population estimation.)
38-43	All law enforcement agencies submitting 12 months of complete arrest data for 2006.	Population estimates for 2006 are based on the percent change in state population from the U.S. Census Bureau's 2005 revised estimates and 2006 provisional estimates. (See the Population estimation.)	The totals provided in Table 43 reflect only those persons arrested for which law enforcement agencies provided race information to the UCR Program; therefore, the totals may not match those shown in other arrest tables for the Nation.
44, 45	All city law enforcement agencies submitting 12 months of complete arrest data for both 2005 and 2006.	The 2006 city arrest trends represent the percent differences between 2005 and 2006 arrest volumes aggregated from all city agencies that submitted complete arrest data for both years. City agencies are all agencies within Population Groups I-VI. (See the Area Definitions.) The population estimates for 2006 are based on 2005 revised estimates and 2006 provisional estimates. Population estimates for 2005 are based on the percent change in state population from the U.S. Census Bureau's 2004 revised estimates and 2005 provisional estimates. (See the Population estimation.)
46-49	All city law enforcement agencies submitting 12 months of complete arrest data for 2006.	City agencies are all agencies within Population Groups I-VI. (See the Area Definitions.) Population estimates for 2006 are based on the percent change in state population from the U.S. Census Bureau's 2005 revised estimates and 2006 provisional estimates. (See the Population estimation.)	The totals provided in Table 49 reflect only those persons arrested for which law enforcement agencies provided race information to the UCR Program; therefore, the totals may not match those shown in other arrest tables for the Nation's cities.
50, 51	All metropolitan county law enforcement agencies submitting 12 months of complete arrest data for both 2005 and 2006.	The 2006 metropolitan county arrest trends represent percent differences between 2005 and 2006 volumes aggregated from contributing agencies. The Metropolitan Counties classification encompasses jurisdictions covered by noncity law enforcement agencies located within currently designated MSAs. (See the Area Definitions.) The population estimates for 2006 are based on the percent change in state population from the U.S. Census Bureau's 2005 revised estimates and 2006 provisional estimates. Population estimates for 2005 are based on the percent change in state population from the U.S. Census Bureau's 2004 revised estimates and 2005 provisional estimates. (See the Population estimation.)
52-55	All metropolitan county law enforcement agencies submitting 12 months of complete arrest data for 2006.	The Metropolitan Counties classification encompasses jurisdictions covered by noncity law enforcement agencies located within currently designated MSAs. (See the Area Definitions.) Population estimates for 2006 are based on the percent change in state population from the U.S. Census Bureau's 2005 revised estimates and 2006 provisional estimates. (See the Population section.)	The totals provided in Table 55 reflect only those persons arrested for which law enforcement agencies provided race information to the UCR Program; therefore, the totals may not match those shown in other arrest tables for metropolitan counties.
56, 57	All nonmetropolitan county law enforcement agencies submitting 12 months of complete arrest data for both 2005 and 2006.	The 2006 nonmetropolitan county arrest trends represent percent differences between 2005 and 2006 arrest volumes aggregated from contributing agencies. The Nonmetropolitan Counties classification encompasses jurisdictions covered by noncity agencies located outside currently designated MSAs. (See the Area Definitions.) Population statistics for 2006 represent estimates based on the percent change in state population from the U.S. Census Bureau's 2005 revised estimates and 2006 provisional estimates. Population statistics for 2005 represent estimates based on the percent change in state population from the U.S. Census Bureau's 2004 revised estimates and 2005 provisional estimates. (See the Population estimation.)
58-61	All nonmetropolitan county law enforcement agencies submitting 12 months of complete arrest data for 2006.	The Nonmetropolitan Counties classification encompasses jurisdictions covered by noncity agencies located outside currently designated MSAs. (See the Area Definitions.) Population estimates for 2006 are based on the percent change in state population from the U.S. Census Bureau's 2005 revised estimates and 2006 provisional estimates. (See the Population estimation.)
62, 63	All suburban area law enforcement agencies submitting 12 months of complete arrest data for both 2005 and 2006.	The 2006 suburban area arrest trends represent percent differences between 2005 and 2006 arrest volumes aggregated from contributing agencies. The classification Suburban Area includes all law enforcement agencies located within a currently designated MSA excluding those agencies that cover principal cities as defined by the U. S. Office of Management and Budget. (See the Area Definitions.) The population estimates for 2006 are based on the percent change in state population from the U.S. Census Bureau's 2005 revised estimate and 2006 provisional estimates. Population estimates for 2005 are based on the percent change in state population from the U.S. Census Bureau's 2004 revised estimates and 2005 provisional estimates. (See the Population estimation.)
64-67	All suburban area law enforcement agencies submitting 12 months of complete arrest data for 2006.	The classification Suburban Area includes all law enforcement agencies located within a currently designated MSA excluding those agencies that cover principal cities as defined by the Office of Management and Budget. (See the Area Definitions.) Population estimates for 2006 are based on the percent change in state population from the U.S. Census Bureau's 2005 revised estimates and 2006 provisional estimates. (See the Population estimation.)	The totals provided in Table 67 reflect only those persons arrested for which law enforcement agencies provided race information to the UCR Program; therefore, the totals may not match those shown in other arrest tables for suburban areas.
68	All law enforcement agencies submitting 12 months of complete disposition data for 2006.	Population estimates for 2006 are based on the percent change in state population from the U.S. Census Bureau's 2005 revised estimates and 2006 provisional estimates. (See the Population estimation.)	In constructing this table, the FBI accepts each individual state's age definition for juveniles.
69	All law enforcement agencies submitting 12 months of complete arrest data for 2006.	Arrest totals are aggregates of the totals reported by individual participating agencies within each state. Population estimates reflect the U.S. Census Bureau's 2006 provisional estimates. (See the Population estimation.)	Any comparison of statistics should take into consideration variances in arrest practices, particularly for Part II crimes. (Offenses in Uniform Crime Reporting defines the UCR Program's Part II offenses.) The Arrest Data section discusses issues affecting specific states in 2006.