Reliability and Changes to FDLE’s Data Collection Method from 1971 – Present
(Note: The data contained in data files available from this website may include updates from participating agencies that are not reflected in the various published reports; therefore, the figures in the two may not match. A discrepancy between published reports and data contained in the data files does not necessarily indicate an error.)
Reliability of Uniform Crime Report (UCR) data is a constant issue in criminology. There are two principle instruments for measuring crime in the US: the UCR and the National Crime Victimization Survey. The consensus in the field is that the "real" amount of criminal activity lies between these two measures.
UCR numbers reflect the crimes reported by the local agencies (primarily Sheriff Offices and Police Departments) to FDLE. The UCR does not include all offenses reported to the police, but is limited to a well-defined list of reportable offenses. These offenses provide an indicator over time of variations in crime trends. In addition, a number of factors influence the reporting of offense incidents to local agencies. For example, some communities are more likely to report a crime to the police than others are. Other factors may include local report-writing policy, manpower allocations, training received by officers on report writing, training received by police records personnel on UCR standards and the decisions and discretion exercised by individuals at every step of the process. As you can see, there are many reasons for variability in reporting between jurisdictions, counties and even states.
Again, UCR is not reporting total crime, but, rather, a select list of crimes reported to the police. This makes the trend data possibly more useful then the actual numbers themselves. It is generally thought that the UCR does a good job of reflecting whether crime is increasing or decreasing. Using the trend, one assumes that any problems in the reporting are consistent over the years even as the problems vary. Nationally, the victimization data (based on interviews of individuals) has mirrored the UCR data trend, which gives us some confidence in its reliability. Simply put, UCR should be used as an indicator of criminal activity but not the ultimate measure.
Additionally, users are cautioned against comparing statistical data of individual reporting units from cities, counties, metropolitan areas, or colleges and universities solely on the basis of their population coverage or student enrollment.
DATA COLLECTION 1971 - Present:
I. Offense Data
In 1971, FDLE began the systematic collection of monthly crime summaries from county and municipal law enforcement agencies in the following crime categories: murder, non-traffic manslaughter, forcible rape, robbery, assault, burglary, larceny, motor vehicle theft, and arson. Over the years several minor modifications were made to the UCR program in response to requests from local agencies and the FBI.
In 1976, Florida was one of the first states to abandon the traditional method of collecting UCR data on a monthly basis and initiated a sophisticated offense-by-offense or incident-based reporting program. Municipal police and county sheriff’s departments reported the individual offense data known to them in the following crime categories: murder, non-traffic manslaughter, forcible rape, robbery, assault, burglary, larceny, motor vehicle theft, and arson. In addition, local law enforcement agencies reported supplemental offense information such as the number of victims, victim’s age, sex and relationship to offender, type of weapon used, the value of property stolen and recovered, circumstances surrounding homicides and other pertinent crime data.
1988 was a transition year for FDLE. Because of a number of changes for the enhancement and expansion of the kind of data collected by the UCR program, complete data are not available for 1988 as some agencies were unable to modify their reporting systems in time to comply with the new format.
1989 represents the first full year of data under the enhanced UCR program. Florida Index crimes were modified to include the new Forcible Sex Offenses category that included forcible rape, forcible sodomy, and forcible fondling. Prior to 1989 forcible sodomy and forcible fondling were reported under aggravated assault, so comparisons to previous years’ Index volumes and rate are to be made with caution. Due to the changes in the Florida Index crimes, the 1989 information can be used as the base for time series comparisons of subsequent UCR data. In addition, Part II offenses became available in this year. Prior to 1989, only arrests for Part II crimes were reported. Part II mandatory offenses include negligent manslaughter, kidnapping/abduction, arson, simple assault, drug/narcotic offenses, bribery, embezzlement, and fraud offenses.
In 1996, the Florida UCR changed its data collection method from a monthly reporting of crime incidents to a semi-annual reporting of summary offense data. With the change from an incident based system to a summary based system, supplementary data such as victim information are no longer available on all offense data except homicide. In addition, with the exception of simple assault, arson and manslaughter, Part II offense data are no longer collected. Due to the changes in the method of data collection, researchers should also be cautioned about comparing data available prior to 1996 to subsequent years.
In 2013, the Federal Bureau of Investigation’s (FBI) UCR Program implemented a new definition of Rape that includes incidents previously reported as Forcible Sodomy. Beginning with the 2013 data, the Florida UCR modified the collection of Forcible Rape data to include forcible rape, attempted rape, and forcible sodomy. When Forcible Rape is presented with a comparison to 2012 data, adjustments have been made for comparison purposes only; the original data presented remains. Caution must be used when comparing Forcible Rape to years prior to 2013.
II. Domestic Violence Data:
Florida began collecting domestic violence data in 1992 for the follow offenses: murder, manslaughter, forcible rape, forcible sodomy, forcible fondling, aggravated assault, arson, simple assault, intimidation and weapons violations. Supplementary victim data was also collected.
In 1996, arson and weapons violations were dropped from the domestic violence count while aggravated stalking and simple stalking were added to the count. Supplementary data are still available for domestic violence with changes in victim data.
III. Arrest Data:
Unlike offenses data, arrest data on Part II mandatory crimes had been collected since 1971. Part I and Part II arrest data include murder, forcible rape, robbery, aggravated assault, burglary, larceny, motor vehicle theft, manslaughter, kidnap/abduction, arson, simple assault, drug sales, drug possession, bribery, embezzlement, fraud, counterfeit/forgery, extortion/blackmail, intimidation, prostitution/commercialized sex, non-forcible sex offenses, stolen property (buy/receive/possess), driving under the influence, destruction/damage/vandalism, gambling, weapons violations, liquor law violations and miscellaneous. Additionally, reports provided information concerning the age, sex, race and residency of all persons arrested by each individual law enforcement agency.
Like with offense data, some agencies were not able to comply with the new format in time, so complete arrest data for 1988 is not available.
Arrest data were modified to include the index offenses of forcible fondling and forcible sodomy.
1996 – 1997
In 1996 and 1997 arrest data were not collected through the UCR system, but through FDLE’s Computerized Criminal History (CCH) database1
. 1996 – 1997 marks a break in the time series arrest data. Due to the different data sources, it is not advisable to compare 1996-1997 arrest data with prior or subsequent years. Arrests in CCH include murder, manslaughter, forcible rape, robbery, aggravated assault, burglary, larceny, motor vehicle theft, other assaults, arson, forgery/counterfeiting, fraud, embezzlement, stolen property, vandalism, weapons violations, prostitution/commercialized vice, sex offenses, sale of Opium/Cocaine, sale of Marijuana, sale of synthetic narcotics, sale of other dangerous drugs, possession of Opium/Cocaine, possession of Marijuana, possession of synthetic drugs, possession of other drugs, bookmaking, number/lottery, all other gambling, offense against family, DUI, liquor laws violations, drunkenness, disorderly conduct, vagrancy, suspicion, curfew/loitering, runaway and all other offenses.
Since the CCH database is constantly updated as new arrest cards are entered, when the 1996 data were pulled for the UCR, 66,067 criminal fingerprint cards had not yet been broken down into the various reporting categories for statistical comparisons. When the 1997 data were pulled from CCH, 118,244 criminal fingerprint cards had not yet been broken down into the various reporting categories. Caution should be used in comparing the individual breakdown categories in that each could be significantly impacted by the records not included. All of FDLE’s arrest data from 1996-1997 is for adult arrests only. Juvenile arrest data was collected from the Department of Juvenile Justice.
1998 – present
In 1998, FDLE went back to collecting arrest data through the UCR system used prior to 1996.
Download the UCR Guide Manual
(Revised December 2013) and the UCR Offense Reporting Hierarchy
for additional information. The UCR Report Forms
are also available.
The CCH database is the State’s central repository for arrest records. CCH data is obtained from arrest fingerprint cards sent to FDLE from local law enforcement agencies.