The Failures Before the Fires: How We Reported This Story
To report this series, the BGA and the Tribune first identified every fatal fire that occurred in a residential building in Chicago from 2014 through 2019.
Although the Chicago Fire Department keeps an annual report of fire deaths, reporters found additional fatalities by reviewing records from the Office of the Illinois State Fire Marshal as well as searching records from the Cook County medical examiner’s office for Chicago deaths that included the words “house fire” in the descriptions of the manner of death.
To be included in the investigation, injuries from the fire must have been listed as a primary or secondary cause of death and the fire must have taken place inside a structure used as a living space. Fatal fires in vacant buildings that were supposed to be secured or demolished to keep people out were also included.
The Tribune and the BGA excluded fires that took place in commercial buildings, vehicles or other structures not used as living spaces. Others were excluded because the victims died before the fires. For example, one fire was excluded because a woman died of a heart attack while smoking a cigarette, which then started the blaze.
In all, 140 fires fit the criteria over the six-year period examined, accounting for 170 deaths.
To ascertain whether there were safety problems in the buildings at the time of the fires, reporters requested fire and police reports for each incident. In addition, the Tribune and the BGA requested records from Chicago’s Department of Buildings showing all violations, inspections and complaints relating to each property going back at least five years before the fire.
In some cases reporters also reviewed court records from civil lawsuits, medical examiner’s reports and news coverage. Reporters also conducted interviews with survivors, relatives and homeowners.
While reviewing the initial records, the reporters’ goal was to identify the number of cases in which city officials had prior warning of fire safety issues that later started fatal fires or otherwise hindered the victims’ chances of survival. They excluded fires where serious fire safety problems played a role in the deaths but there was no documented evidence that city officials were aware of the problems.
In the end, 42 of the 140 fatal fires met the criteria.
Documenting fire safety issues
There were two main ways reporters confirmed fire safety issues existed in the buildings where fatal fires occurred. One was if firefighters reported that an issue (such as faulty wiring or a lack of heat) had definitely or possibly caused the blaze. The second was if interviews or records about the fire revealed safety issues that could affect chances of survival, regardless of how the blazes began.
Issues in the second group included blocked or insufficient exits, overcrowded or illegal units, a lack of functioning smoke alarms (whether they were not working or missing altogether) and, for older high-rises without sprinkler systems, violations of the building’s Life Safety Evaluation. In most cases, records about the fires showed the fire safety issues directly hampered residents’ chances of survival. In a few cases, records did not clearly document whether the fire safety issues directly hampered residents or firefighters’ response during the fire.
To establish the causes of the fires, reporters reviewed investigation reports by firefighters and police.
Sometimes, investigators reached no definite conclusion about the cause. Often, investigators listed two possible causes instead. For example, firefighters might report finding an electrical device near the area where the fire began but then state an electrical engineer would be needed to determine whether it caused the blaze. Many cases were closed with a finding that an “open flame” had ignited available materials and offered no further detail.
In some cases, fire and police concluded the fires were accidental in nature and closed the investigation because there was no criminal wrongdoing, the question at the heart of these investigations. Some fires created such severe damage that it was unsafe or impossible for officials to assess the scene.
Determining prior city knowledge
To determine whether city officials had prior knowledge of the relevant fire safety issues at these properties, reporters reviewed complaints filed to the 311 hotline and inspection reports from the Department of Buildings.
Complaints to 311 are unverified reports of violations. In some cases, the city conducted an inspection and documented whether the problem did or did not exist. In others, there is no record the city conducted an inspection. Reporters considered the existence of a complaint to be proof the city was notified of a problem if other records confirmed the problem existed at the time of the fire.
Reporters did not focus on whether the exact fire safety violation documented in a complaint or inspection also was cited in a fire or police investigation. One reason is that the location of the violation — such as a specific apartment — was often redacted in the records provided by the city. Inspectors also often reported not being able to access all parts of a building, making it impossible to know if a problem found in one area also existed elsewhere. If the property spanned multiple street addresses, as is common for courtyard buildings, reporters considered violations at any of those addresses.
Reporters included the case if the general fire safety issue was found to play a role in the fire. An example would be if building inspectors found missing smoke detectors in a common area of an apartment complex and firefighters also reported smoke detectors were missing or not working in a different area affected by the fire.
For each of the 42 cases with evidence the city had prior warning of fire issues, reporters created a database of complaints related to the properties received by the Department of Buildings between 2009 and the dates of the fires. Reporters hand-entered the information in each complaint into a spreadsheet, noting the date of the request, a description of the problem and the resolution of the complaint.
In cases where inspectors documented problems, there are three options for enforcement action. The least stringent is to send a warning letter to the property owner noting the deficiencies. This option isn’t tied to any reinspections or fines, and reporters did not consider a warning letter to be evidence a problem was solved.
The other two options are to send violations to the Department of Administrative Hearings or to file a case in Circuit Court, which is handled by the Department of Law. When violations ended up in an administrative hearing or court, reporters requested all records related to the proceedings. Reporters considered the violations to be unaddressed if the city provided no records showing inspectors had verified repairs in person before the fire.
Assessing the presence of smoke detectors
In Chicago, the law holds landlords responsible for installing working smoke detectors in common areas and providing them for tenants’ use in individual units. Tenants are in charge of changing batteries as needed. However, property owners can be cited if inspectors find smoke detectors with no batteries inside tenants’ units.
Reporters assessed the presence of working smoke detectors in each of the 140 fatal fires. Often, the fire reports provided incomplete or conflicting information. For example, in one report, officials recorded both “detector presence: N - non present” and “smoke detectors: present.” In the same report, firefighters stated they did not hear any smoke detectors while responding to the blaze.
To deal with these discrepancies, reporters reviewed other available information for each case, giving more weight to statements about smoke detectors from people referenced in official reports or interviewed by reporters. In the case above, for example, reporters ultimately concluded there were no working smoke detectors because firefighters said they did not hear alarms.
In cases where officials reported the building contained some working smoke detectors but others were missing or not working, reporters recorded the presence of smoke detectors as “partial.”
Analyzing clearance rates
The Tribune and the BGA sent a public-records request to the Department of Buildings for data on “serious” violations citywide from 2014 through 2019. The definition of “serious” comes from city rules established to identify problem landlords. Reporters used this dataset to analyze the rate at which inspectors marked violations as fixed.
The dataset included address and ZIP code, ward, violation code, status, inspection type and supervisor recommendation. The status of the violations was either closed or open; recommendations were to send a notice, refer the violation to the Department of Administrative Hearings or file a lawsuit in Cook County Circuit Court.
The Tribune and the BGA also used the dataset to compare mainly white areas of the city with mainly non-white areas on rate of complied violations. To do this, reporters joined the ZIP code data from the city to data on race and ethnicity provided by the U.S. Census Bureau by ZIP Code Tabulation Area. A ZIP code was considered to be majority white if more than 50% of people living there reported they were white and non-Hispanic.
Administrative Hearings analysis
To assess the effectiveness of the Department of Administrative Hearings at ensuring building violations are fixed, the BGA and Tribune conducted an in-depth examination of 50 administrative hearings cases related to buildings affected by the 42 fires. Some of the properties accounted for multiple administrative hearings cases in the review, and other buildings had none. The violations in these cases were not exclusively related to fire safety.
For the 50 cases, reporters filed public-records requests for the complete case file, including audio recordings from each hearing date. They then reviewed all of the records to determine the number and type of violations at issue, the number of hearing dates and the reasons given for each extension, any reinspection orders or attempts, and the final status of each violation. This included information regarding whether and how the city verified that the property owner had fixed the violations.
Reporters considered cases closed without repairs when an owner admitted to not fixing one or more problems and the city closed the case anyway. Sometimes this occurred because landlords chose to pay fines instead, or city lawyers agreed to disregard some violations as long as the property owner fixed the most serious problems. In other cases, it is unclear why the city closed cases before problems were fixed.
The BGA/Tribune review found it is common practice for hearing officers to accept landlords’ evidence — including photos, store receipts and statements solicited from tenants — as proof of the required repairs without verifying the information through in-person inspections. The BGA and the Tribune did not consider landlords’ evidence as proof a fire safety hazard was fixed.
The problem landlord list
A 2015 ordinance created a list of “bad” or “problem” landlords who would be barred from most city contracts until violations at their buildings were fixed. City officials published the list only twice before it was abandoned the following year.
A total of 44 buildings and their landlords ended up on the list before the city stopped using it.
To examine how many more buildings could have qualified had the city continued enforcing the list, reporters started by reviewing a public dataset on administrative hearings cases and their outcomes.
Under the ordinance, a building and its landlord was eligible for the list if:
◘ The building was the subject of at least two administrative hearings cases within two years.
◘ In each of those cases, hearing officers determined the owners were responsible, or “liable,” for at least one violation.
◘ Out of all violations included in the two or more cases, there were at least three “serious” problems, as defined in the city’s rules.
◘ One of the findings of liability was for a “serious” violation.
While the city says it used the same public dataset reporters analyzed to help determine whether buildings were eligible for the problem landlord list, it is impossible to identify all qualifying buildings based on that information alone.
That’s because administrative hearings cases often include more than one violation, and records reviewed by the Tribune and BGA show outcomes for each issue often vary. However, the dataset doesn’t include those individual outcomes, instead showing only a general disposition for the entire case.
For example, one administrative hearings case may contain five violations. If one of those violations was closed out with a finding of “liable” but the others were dismissed, the public dataset would show a finding of “liable” for the case. From this information, it’s impossible to tell how many violations were substantiated.
The Tribune and the BGA did use the data to estimate how many buildings could have qualified for the list.
Reporters started by identifying buildings where more than one administrative hearings case was marked as “liable” within a two-year period. Then, reporters eliminated sets of cases with fewer than three serious violations in total.
This review showed hundreds of buildings fulfilled the first three qualifications in the city’s rules. Only a review of the individual case files could determine the number of buildings that meet the final criterion, a finding of “liable” for at least one serious violation.