unsaved:///newpage2.htm

REPORT CONCLUSIONS

On December 15, 1997, the Office of the State Fire Marshal (OSFM) merged arson investigators into the Fire Prevention Division and established a protocol which specified OSFM’s policy for responding to requests for arson investigators. These changes were intended to improve the accountability of investigators who work from home. The results of the audit were as follows:

The published protocol stated investigators would respond immediately to fires involving a fatality while other fires would normally receive a response during daylight. The prior OSFM policy was to investigate nearly all types of fires immediately. Most other states’ fire marshals and Illinois’ fire chiefs said in our survey their investigators respond immediately or as soon as possible to most fires.

In the fire cases we randomly sampled, OSFM overall response was quicker prior to the establishment of the protocol.

Controls over the arson investigation program were strengthened after the December 1997 reorganization and resulted in less overtime and cases being closed faster.

OSFM needs to improve controls further. The number of investigations conducted annually varied significantly by investigator from 8 to 91 investigations. Many fiscal year 1998 fire investigations sampled lacked physical evidence (37%), interviews (24%), or reports to police (18%), although there was improvement after the 1997 reorganization. Investigation reports were not reviewed by supervisors and the policy to have investigators review each others’ reports was not complied with in almost half the cases sampled.

The arson investigation program lacked adequate supervision after the division director and supervisors were laid off in May 1995 for budgetary reasons. The program was assigned supervisors after the 1997 reorganization but supervisors were not certified arson investigators as they were before the 1995 layoffs. Two-thirds of other states’ fire marshals and three-fourths of Illinois’ fire chiefs who responded to our survey questionnaire said their supervisors were certified investigators.

The amount of accumulated compensatory time increased significantly from 104 hours on June 30, 1995 to 2,452 hours on December 15, 1997. After the reorganization, accumulated compensatory time declined to 2,189 hours on July 1, 1998.

FIRE MARSHAL’S ROLE

The mission of the Office of the State Fire Marshal is to reduce death, injury, and property loss from fires and other hazards. When requested, OSFM assists the local fire service determine the origin and cause of fires, and may help apprehend suspected arsonists. This service is requested mainly by rural or volunteer fire departments who lack trained investigators.

The arson investigation program is governed by the Fire Investigation Act (425 ILCS 25/6) which places investigation responsibility on local fire chiefs but gives OSFM authority over such investigations if OSFM deems it necessary or expedient. OSFM has established a protocol as its criteria for determining if a fire investigation is necessary or expedient.

OSFM had 19 arson investigators in fiscal year 1998 who worked from their homes which was their duty station. They responded to 1,017 fire cases during fiscal year 1998 which were categorized as follows (see Digest Exhibit 1):

456 incendiary or arson (45%)

184 accidental (18%)

273 undetermined (27%)

104 requests for canine (K-9) only (10%)

Before May 1995, there was a separate division for arson investigation but due to a budget shortfall, the arson division director and all supervisor positions were eliminated. From May 1995 to December 1997, arson investigators reported directly to the agency’s deputy chief executive officer and were organized in four regions: Chicago area, Northern, Central, and Southern areas of Illinois.

In December 1997, the agency moved arson investigators into the Fire Prevention Division and reorganized it as the Fire Prevention and Arson Investigation Division. The Fire Prevention Division had three regions headed by a Regional Administrator (supervisor) and arson investigators began reporting to these three Regional Administrators. Agency officials said investigators were assigned to 18 geographic districts, instead of four regions, to improve control. (See report pages 2–7)

PROTOCOL ESTABLISHED

On December 15, 1997, the Office of the State Fire Marshal established a protocol which sets priorities for arson investigators to respond to fires. The protocol was published in an OSFM newsletter and has four codes which specify OSFM’s response priorities for dispatching arson investigators (see Digest Exhibit 2).

Digest Exhibit 2
Protocol
December 15, 1997

Code 1 is for fatality and is stated to receive an immediate or quick response.
Code 2 is for injuries and is stated to receive a response during daylight.
Code 3 is for suspicious fires and is stated to receive a response during daylight.
Code 4 is for smaller fires, such as car or garage fires, and is stated to receive a response the next business day.

Source: Office of the State Fire Marshall

In our random sample, only 6 of 60 cases (10%) after the protocol fit the first priority; the remaining 90 percent fit the other three priority codes.

Although the primary objective of the arson investigation program is to investigate suspicious fires, they were made the third priority. Also, most fires occur in the evening or at night.

Most other states’ fire marshals and Illinois’ fire chiefs said in our survey questionnaire their investigators responded immediately to most fires, similar to OSFM’s procedure before the protocol. See Digest Exhibit 3.

Response was faster before protocol.
Controls were strengthened after the 1997 reorganization but need further improvement.
Some investigation files lacked complete documentation
OSFM arson division supervisors were not certified arson investigators as they were in 67% of other states’ fire marshals

Published protocol specified immediate response only for fatalities.
Protocol made suspicious fires, which are the primary responsibility of the arson unit, third priority.

Digest Exhibit 3 COMPARISON OF RESPONSE POLICIES Fiscal Year 1998
			Immediate or ASAP
Code	Type of Fire	OSFM Policy	Other States	Fire Chiefs
1	Fatality	Immediate	92%	90%
2	Injury/Major Economic Impact	Same or next day	73%	90%
3	Suspicious	Same or next day	73%	89%
4	Garage/Car/Barn	Next business day	63%	81%
SOURCE: OAG surveys of fire chiefs and other states’ fire marshals, and OSFM.

75% of the fire chiefs were satisfied with investigations; 30% were dissatisfied with the protocol (response priorities).
Investigators arrived within four hours of the fire in 51% of cases before the protocol vs. 42% after the protocol.
Quick response helps investigators talk to witnesses, take pictures, secure fire scene, and collect evidence.

The protocol made two important changes. First, investigators can no longer be called directly but have to be requested through the Arson Hotline. Officials said this practice of being able to contact investigators directly had resulted in a higher number of calls and accumulation of compensatory time. Second, suspicious fires are not required to be investigated immediately but may be investigated during daylight the same day or the next day. Previously, the written policy was that response to small fires could be delayed until daylight.

In February 1998, OSFM issued a clarification to the protocol which stated supervisors had authority to dispatch investigators at any time. Agency officials noted the clarification was reviewed and approved by the Illinois Fire Chiefs Association.

A quick response may permit investigators to talk with witnesses, take pictures of the fire scene, take pictures of people observing the fire, secure the fire scene, and collect physical evidence.

Although the arson investigation program assists local fire departments, their input was not obtained when the protocol was developed and almost one-third (30%) said in our survey they were dissatisfied with the protocol. When questioned about other aspects of the arson investigation program (i.e., overall satisfaction, adequacy of investigations, and timeliness), over 75 percent said they were satisfied. (See report pages 9–14)

RESPONSE TIME

In our random sample of cases, OSFM investigators responded more quickly to fires before the new protocol. Investigators arrived within four hours of the fire in 51 percent of the cases sampled before the protocol as compared to 42 percent of the cases after the protocol (see Digest Exhibit 4). Calculations in one-hour increments also showed that response was quicker before the protocol.

Median response time before the protocol was 3½ hours from the time of fire to the time investigators arrived, as compared to 5½ hours after the protocol.

We also reviewed the response time after the agency was contacted. The time that OSFM was contacted to send investigators before the protocol was not documented. However, after the protocol, the median response time from the time the agency was contacted to the time investigators arrived was faster during business hours: 1 hour 46 minutes for calls during business hours versus 4 hours 5 minutes for calls after business hours.

Our sample showed that 9 of the 10 longest response times after the protocol were for calls received after business hours, with the two longest response times (over 3½ days) being for calls received on a Friday.

Percentages of responses before protocol does not add to 100% due to rounding. This exhibit includes all cases: investigations and K-9.

Source: OAG sample of OSFM fire investigations cases conducted during fiscal year 1998.

Cases were closed faster: 3 days after protocol vs. 14 days before protocol.
Annual caseload per investigator varied from 8 to 91 investigations in fiscal year 1998.
Average number of investigations was less than 1 per week per investigator.

OSFM supervisors did not review investigation reports.
Reports were reviewed by supervisors at 85% of other states' fire marshals.
Investigation manual was last revised 11 years ago.

Another factor associated with investigations was the number of days that elapsed to close an investigation case. After the protocol, cases were closed much sooner with median completion time being 3 days versus 14 days before the protocol in the cases sampled. (See report pages 15–17)

CASELOAD AND OVERTIME

The number of investigations conducted by OSFM investigators varied significantly in fiscal year 1998 and ranged from 8 to 91 investigations. The average caseload per investigator was under one investigation per week. The differences among investigators’ caseloads indicates the agency needs to review and improve controls to reduce the disparity so that all investigators are utilized more equally and efficiently.

Total accumulated compensatory time grew from 104 hours on June 30, 1995 to 2,452 hours on December 15, 1997. After the reorganization, accumulated compensatory time declined to 2,189 hours on July 1, 1998.

In fiscal year 1998, arson investigators earned, used, and liquidated less compensatory time than during fiscal year 1997. Investigators earned 4,493 hours of compensatory time in fiscal year 1998 as compared to 7,352 hours in fiscal year 1997. They used 3,877 hours and liquidated 40 hours in fiscal year 1998. In fiscal year 1997, investigators used 5,471 hours and liquidated 916 hours. (See report pages 19–25)

SUPERVISION

Between May 1995 and December 15, 1997, there were no arson division supervisors or director; instead investigators reported directly to the agency’s second highest executive officer.

Supervisory personnel were assigned after December 15, 1997; however, none of the investigation reports sampled indicated their review. One reason cited by OSFM officials for not having supervisors review investigation reports was that these reports are subject to subpoena and cross-examination in a court. Also, the division’s supervisory personnel were not certified arson investigators. Before the layoffs in 1995, division supervisors and the director were certified arson investigators and one of the supervisors’ responsibilities in the position description was to review and evaluate ". . . staff’s arson reports for technical accuracy and completeness."

The responsibility for reviewing reports was instead assigned to Arson Investigators II. In our sample, 56 of 120 case reports (47%) were not reviewed by Arson Investigators II. Furthermore, five investigations (4%) sampled did not have reports prepared (see Digest Exhibit 5).

Digest Exhibit 5 CASE RECORDS BEFORE AND AFTER REORGANIZATION Fiscal Year 1998
	Before	After	TOTAL
No interviews conducted	35%	14%	24%
No physical evidence collected	40%	34%	37%
No law enforcement notified for fires whose cause was arson or was undetermined	26%	10%	18%
Case closed same day	23%	23%	23%
No case initiation form *	100%	38%	69%
No Fire Investigation Report *	0%	8%	4%
No report review by an Arson Investigator II * ^A	58%	35%	47%
*Includes K-9 cases for which these records were applicable. A OSFM policy was to have Arson Investigators II review investigation reports. (There were no documented reviews by supervisors.)
SOURCE: OAG sample of 60 fire investigation cases before, and another 60 cases after, the reorganization.

Approximately ten percent of the fire investigation reports sampled did not meet the level of completeness set forth in the arson investigation Operations Manual.

Most other states’ fire marshals (85%) and Illinois’ fire departments (68%) said their supervisors reviewed investigation reports, such as for procedural compliance, technical accuracy, investigative sufficiency, and completeness.

In our survey of other states’ fire marshals, 16 of 24 responding states (67%) said their supervisors were certified fire or arson investigators. In our survey of Illinois local fire chiefs, 110 of 146 responding fire chiefs (75%) said their supervisors were certified fire or arson investigators.

Another item that needs to be reviewed is the arson investigation Operations Manual. The manual was last revised more than a decade ago in 1988 and should be updated to make it current with statutes, methods, and procedures. (See report pages 31–38)