2005 Vermont Crime Report

Introduction

The 2005 edition of the Vermont Crime Report contains an analysis of crime reported to law enforcement agencies throughout the State of Vermont during calendar year 2005. Crimes reported to state police, game wardens, municipal police, sheriff departments, and other state law enforcement agencies are contained in this report. Crime statistics are arranged statewide, by county, and by town.

VERMONT CRIME ON-LINE: A new format for the Crime Report

The State of Vermont's Uniform Crime Reporting program has been in transition for several years. The program has been changing from a summary-based to an incident-based reporting system. The summary-based system which has been in effect since 1984 in Vermont provides information on only a limited number of crimes referred to as Part I and Part II crimes. The incident-based system reports on a much broader range of crimes and includes data on the circumstances of the crime including characteristics of victims and defendants. Vermont's transition mirrors similar changes in other states as law enforcement agencies across the country implement the Federal Bureau of Investigation's new crime reporting system known as the National Incident-Based Reporting System or NIBRS. In Vermont the new NIBRS format is known as Vermont Crime On-Line or VCON. 2004 was the debut year for VCON.

Vermont Crime On-Line provides the full complement of data elements specified by NIBRS. VCON allows the user to view standard reports or to create their own custom reports on-line. Another advantage of the VCON format lies in enhanced data quality – more data and more accurate data. While the summary-based format collected crime data on 25 crimes, VCON collects data on 56 crimes. In addition, VCON reports information on the time and day when crimes occur, crime location type, the amount of property crime loss, victim injuries, statistical information on the age, gender, and race of both victims and defendants and other crime circumstance data of interest to the public, law enforcement, public administrators, legislators, and researchers.

The enhanced accuracy of VCON data stems from the manner in which it is collected. In the summary-based system, crime data is reported directly to VCIC with little or no error checking performed by the reporting agency. The VCON system requires rigorous data quality auditing before the data can be submitted to VCIC for inclusion in the VCON system. Consequently, the incidence of crime in a community will be more accurately recorded due to the automated error checking in the reporting software.

VERMONT CRIME ON-LINE Methodology

What is Included in VCON

The crime data provided for Vermont Crime On-Line is an enumeration of crimes known to law enforcement agencies. It is not an enumeration of all crimes that were committed in Vermont during calendar year 2005. Crimes that are included in VCON are based on reports received by law enforcement agencies from victims, officers who discover infractions, or other sources. Crimes that may have occurred but were not reported are not included in this report. All reports of crime have been validated by a law enforcement officer. That is, reports which are later shown to be unfounded (e.g., property reported as stolen but later discovered as misplaced) are not included in this report.
All agencies were required to submit crime data for Group A and Group B crimes as defined by the Federal Bureau of Investigation's NIBRS program. The definitions for Group A and Group B crimes are accessible from the Vermont Crime On-Line Site.

There are two important differences between the way in which crime is counted by VCON vs. the way crime was counted by the old summary system:

1. VCON collects data on 56 crimes vs. the 26 crimes counted in the old summary system. This change affects the way in which the data is displayed rather than the total amount of crime counted. VCON provides a greater level of detail than the old summary system.

2. VCON only collects data for Group B offenses (a group of 11 less serious crimes) when there is an arrest. The old summary system as implemented in Vermont collected data on all offenses. This change does result in a reduction in the total number of crimes reported on VCON.

The changes in the way in which crime is counted will result in a lower number of crimes reported in VCON than would be the case if the old summary system were used as the basis for the Crime Report.

Data Quality

The data in this report were provided by municipal, sheriff, state police, and other Vermont state law enforcement agencies which are solely responsible for the accuracy of their submissions. Agencies are required to utilize automated editing software to error check their data prior to submitting the information to the Vermont Crime Information Center (VCIC). The FBI performs similar edit checks on Vermont’s data before including the data in the national system. The data which appears in VCON has undergone edit checking at both the local and federal level before being published.
The data quality of the VCON system is higher than that of the old summary system due to automated vs. manual data collection methods and the data quality editing that is performed at both the local and federal level. As such there will be some discrepancies when comparing crime data collected using the summary format vs. data collected using VCON. Due to more rigorous data collection standards the amount of crime reported in VCON will be less than the number of crimes reported in the old summary system.

Location Issues

Municipal police departments generally submit crime data which is limited to their town/city. Sheriff Departments report crime data for towns for which they have law enforcement responsibilities. State Police and other state law enforcement agencies report crime data based on the location of the crime. It is not uncommon for towns to have different law enforcement agencies policing the jurisdiction depending on the time of day or day of the week. In these cases, the Vermont Crime Information Center (VCIC) collates crime reports from all involved law enforcement agencies to ensure the accuracy of the town's crime rate.

Reporting Compliance

In 2005, crime data was received from all law enforcement agencies required to report with the exception of: Bristol Police Department, Lyndonville Police Department, Middlebury Police Department, Addison County Sheriff, Caledonia County Sheriff, Chittenden County Sheriff, Essex County Sheriff, Department of Motor Vehicles, and the Department of Liquor Control. As such, no data from these agencies are included on the Vermont Crime On-Line site.

Population Figures

Population figures provided on the VCON site are based on the most current data available from the Vermont Department of Health.

2005 Report Methodology

As discussed in the Methodology Section above there are discrepancies between crime counts reported in VCON vs. crime counts which would be reported using the old summary system. These discrepancies have arisen for three reasons: 1) changes in the way crimes are counted; 2) more rigorous data quality standards; and 3) the failure of some law enforcement agencies to report.

In light of these discrepancies the 2004 Crime Report was available in both the Summary and VCON formats. The Summary data was provided by reporting agencies directly from their records management systems and was not subject to the more rigorous data quality standards of the VCON program. The VCON data was subject to rigorous data quality editing. When the Summary data was compared to the VCON data discrepancies were apparent as expected – Summary data reported more crime than did VCON data.

In an effort to more fully understand these discrepancies, the Vermont Crime Information Center (VCIC) conducted an audit of the 2005 crime data as it appears in reporting agency record management systems and the 2005 crime data that was reported to VCON. It was determined that data in the reporting agency record management systems was over reported and data in VCON was underreported.

Over Reporting

The audit indicated that approximately 12% of crime reports were not classified correctly. For example a case may have been classified for reporting purposes as a rape when in reality it was a case of forcible fondling. Errors of this sort would result in an over reporting of the number of rapes. In other cases crimes were reported more than once to VCON. For example a robbery may have occurred in Town X and was reported to VCON by the Town X Police Department. Another agency may also have responded to the call and/or assisted in the investigation and reported the same robbery in their records management system as well. The result is that two or more robberies were reported to VCIC when only one actually occurred.

Underreporting

The audit indicated that the data quality of crime information reported to VCON by reporting agencies was excellent. In less than 1% of the cases were classification problems found in the VCON data. It appears that the rigorous data quality auditing that is required by the VCON program is paying off in terms of increased data accuracy. However, the audit did indicate that some crimes that were verified as accurate in reporting agency records management systems were not reported to VCON. Thus though the data that was reported to VCON was of excellent quality it was underreported.

Data Quality Strategy

In light of these findings VCIC has decided to discontinue the use of unaudited data from reporting agency record management systems as the basis for reporting Statewide Crime Statistics. Rather, only crime data reported to VCON and subject to data quality audits will be used to calculate Statewide Crime Statistics. Though this change will increase the accuracy of the data there still remains the issue of underreporting. To resolve this issue reporting agencies were notified in June, 2006 that they would have 60 days to review their 2005 crime data and verify that all valid cases have been sent to VCON. As a result of this effort more than 2,000 additional cases were reported to VCON.

2005 Report Format

For purposes of continuity, the 2005 Crime Report does contain a Statewide Crime Statistics table but the data is based only on the Crime Index Crimes (known as Part I crimes) designated by the FBI. The data in this table were obtained from VCON and collapsed into the Summary system crime types. In the past the Statewide Crime Statistics table contained both the Crime Index Crimes and a list of less serious crimes known at Part II crimes. Because of different reporting practices between the Summary and VCON system it is not possible to collapse VCON data into a valid Summary, Part II crime table. A complete enumeration of crimes reported to law enforcement is available from VCON. Crime statistics are arranged statewide, by county, and by town.

Because the data in the 2005 Crime Index table is based exclusively on audited data from VCON, the data in the 2005 Crime Index table indicates a lower volume of crime when compared with the Statewide Crime Index table (Part 1 crimes) from earlier years. Though VCIC is confident that the crime data reported in the 2005 VCON Vermont Crime Report is more accurate then the data in past years, the question arises as to whether the lower volume indicates a reduction in crime or is the lower volume a function of the change in the data collection method?

Implications

By estimating the 2005 Crime Index using the summary-based data collection method used in 2004 there appears to be a 1% - 3% reduction in Crime Index crimes between 2004 and 2005. Homicides were down from 11 victims in 2004 to 8 victims in 2005. Though homicides were down, the Violent Crime Index (Homicide, Rape, Aggravated Assault and Robbery) collectively increased by approximately 2.5%. The increase was due to an increase in Aggravated Assaults. The Property Crime Index (Burglary, Larceny, Motor Vehicle Theft, and Arson) declined by 3%. These findings are consistent with figures released by the FBI in June which indicated that the Violent Crime Index increased nationally by 2.5% and the Propertry Crime Index declined by 1.6%.