Fire Safety Inspections in San Francisco a Tale of Two Departments:

Considerable Resources Are Wasted Creating R­2 Lists Because HIS needs to know the last inspection date for R­2s in order to identify which R­2s are due for a routine inspection, and CTS cannot generate an accurate report containing this information, HIS had to create a “work around” by manually preparing R­2 lists that included accurate last inspection dates. The process for creating the R­2 lists begins with getting the list of R­2s in the City from DBI MIS. HIS cannot get an accurate list of all the R­2s in the City without the help of DBI MIS. We ​ were told that there is one Oracle database that stores information on all the residential buildings in our City, including the property address, property owner contact information and some building characteristics, such as the number of units in each building. This Oracle database is not integrated with any other City department database­­including CTS. HIS does not have access to this database; however, DBI MIS does. Thus, HIS must ask DBI MIS to generate an initial R­2 list that includes all residential buildings with three or more units (R­2s) that are located in the specific area(s) of the City in which HIS will be conducting routine inspections. The initial R­2 list includes the property addresses and contact information for the the property owners. However, it does not list any routine inspection dates. The list of R­2s targeted for routine inspections is created manually. After HIS receives the ​ initial R­2 list from DBI MIS, support staff or inspectors must identify the last routine inspection date by looking up this information for each property, one property at a time, in CTS. Thereafter, the last inspection dates are added to the Excel spreadsheet, which can then be sorted by last inspection date, and the R­2s that are due for a routine inspection can be easily identified. DBI MIS did not generate the R­2 lists for the first six rounds of Focused Code Enforcement. As a result of the series of fires that occurred in the Mission starting in late 2014 ​ (See Introduction), the Board of Supervisors, along with other government officials, made Fire Safety Inspections in San Francisco 13 inquiries into the causes of these numerous fires and asked how San Franciscans could be better protected from harm, property loss and displacement as a result of fires. In response, HIS beefed up its code enforcement (including fire safety) by assembling a team of inspectors to conduct a blitz of routine inspections along major corridors in the City. This began shortly after the catastrophic fire at 22nd and Mission Streets. HIS refers to this program as “Focused Code Enforcement.” This program was a huge departure from how routine inspections had been conducted in the past when all HIS inspectors were assigned to specific districts and required to investigate R­2 complaints in those districts, in addition to conducting routine inspections throughout San Francisco. There were no inspectors dedicated to working exclusively on routine inspections during this time. Sometimes, these routine inspections were performed at opposite ends of the City from each other and nowhere near the inspectors’ districts. We were told that investigating complaints was prioritized over conducting routine inspections back then. A few years ago, a couple of inspectors were taken out of districts and assigned to work strictly on routine inspections. After that, Focused Code Enforcement became the model for conducting routine inspections. We were told that DBI MIS did not create the initial R­2 lists for HIS during the first six rounds of Focused Code Enforcement because DBI MIS was fully committed to the Accela project, a proposed new computer system. (See CTS Is Outdated, below.) Instead, during that time, an inspector volunteered to create the initial list of R­2s for focused code enforcement himself. He did this by combining an old Excel spreadsheet that listed R­2s in districts that were covered by another inspector with his own personally developed list of R­2s located in the districts that he covered. Since the property owner contact information for the R­2s may have been outdated on the initial R­2 list he created, he then had to go into CTS and look up current property owner contact information, one property at a time, for each R­2. After that, he had to go to a different screen in CTS to look up each R­2’s last inspection date. We were told that this “work around” was very labor intensive. The first round of focused routine inspections was conducted along the Mission Street Corridor (along Mission Street starting at the Embarcadero south to where Mission turns into Daly City). Subsequent rounds were performed in targeted areas of the Mission, Chinatown, North Beach, the Marina, Pacific Heights, Inner Richmond, Outer Richmond and the Tenderloin. According to interviews conducted with DBI staff, these areas were chosen because they included many R­2s with high risk characteristics for fire­­older wood buildings that contain both residential and commercial units (that may also have tenant overcrowding) and are situated along congested commercial corridors. The focused routine inspections conducted along the Mission Street Corridor, in the Mission, Chinatown and the Tenderloin were more extensive (included more buildings) than the focused routine inspections conducted in the other five districts listed above. We were told HIS does not have enough inspectors to conduct focused routine inspections on all the R­2s due for a routine ​​​ inspection in these districts. Fire Safety Inspections in San Francisco 14 HIS will soon begin a seventh round of focused routine inspections targeting 200 R­2s throughout San Francisco. We learned that DBI MIS helped create the R­2 list for this seventh round of Focused Code Enforcement.

CTS Reports Do Not Capture The Various Ways Routine Inspections Are Documented When HIS inspectors conduct routine inspections, they inspect the common areas of R­2s for 35 health and safety items. Fifteen of these items are fire safety related. (See Appendix, Exhibit 6.) ​​ There are two types of routine inspections­­scheduled routine inspections and complaint­generated routine inspections. Scheduled routine inspections are scheduled in advance with the property owner and are not performed in connection with any other inspection. Typically, they are conducted by inspectors who focus on routine inspections. We will refer to these scheduled routine inspections simply as routine inspections. ​ Inspectors do not always choose the same “Source” for routine inspections. Inspectors ​ document their routine inspections and complaint investigations in two screens in CTS­­the complaint screen and details screen. (See Appendix, Exhibits 7 and 8.) On the complaint screen, inspectors enter basic information such as their name, the R­2’s address, the date of the inspection and a narrative describing the inspection in the “Description” box. They also document the “Source” from a drop­down menu with 33 options. (See Appendix, Exhibit 9.) The Source serves a dual purpose of capturing either who referred the inspection to HIS or the type of inspection that was performed, such as “Routines” or “Complaint.” Based on our interviews, we learned that inspectors do not always choose the same Source for documenting routine inspections. Most inspectors will choose “Routines” as the Source for routine inspections; however, some will choose “Routine Appointment Letter.” In the case of 2960 California Street, the inspector chose “Telephone” as the Source for the routine inspection he conducted on September 6, 2010. (See Appendix, Exhibit 4.) Since Telephone is not typically used as a Source for routine inspections it fell outside the parameters of the report we asked DBI MIS to generate for last routine inspection dates. We do not know if Telephone was erroneously chosen or if there was a legitimate reason for documenting the routine inspection that way. Regardless, it serves as an important example of a CTS report not meeting the needs of the end user because the report parameters were not adequately defined and agreed upon by the both the report generator and end­user beforehand. Inspectors do not always choose the same “Abatement Type” for the initial routine inspection. The “Abatement Type” is meant to document the action the inspector took. There ​ are 62 choices on the Abatement Type drop­down menu. (See Appendix, Exhibit 10.) Although most inspectors told us that they choose “Inspection of Premises Made” for routine inspections, some inspectors choose “Case Received.” From our interviews with HIS inspectors and DBI MIS and seeing many of the standard reports that are available in CTS but not used by HIS, we have concluded that CTS report parameters are not adequately defined. We think responsibility for this rests with HIS and DBI MIS. Fire Safety Inspections in San Francisco 15 3. Complaint­Generated Routine Inspections Are Not Always Documented in a Way That Can Be Measured by CTS Scheduled routine inspections and complaint­generated routine inspections (“CG routine inspections”) are conducted differently. As a result, the way they are documented differs. In contrast to routine inspections, CG routine inspections are not scheduled in advance and are conducted while a district inspector is already at an R­2 investigating a complaint. Rather than having another inspector go back to the R­2 to conduct a separate routine inspection, district inspectors have been instructed to conduct a CG routine inspection while they are already at an R­2. A complaint investigation focuses on the complaint, for example “heat not working at my unit,” and will not include the common areas of an R­2 unless it is the subject of a complaint. District inspectors schedule the complaint investigation with the complainant (usually a tenant). However, the CG routine inspection is not scheduled with the property owner. Instead, the CG routine inspection can be conducted without prior scheduling because once the district inspector has been let into an R­2 to investigate a complaint, he will have access to the common areas of the R­2 and can conduct a routine inspection. Since the routine inspection arises from the complaint investigation, they are called complaint­generated routine inspections. District inspectors do not always choose the same “Source” for the CG routine inspection. District inspectors primarily investigate R­2 complaints in their districts. They also conduct CG routine inspections. They must document both. When documenting a complaint investigation, the Source is “Complaint.” However, there is no specific Source for CG routine inspections. We were told that that there used to be “Complaint Generated Routine” listed on the Source drop­down menu but it was taken out a few years ago. As a result, inspectors document their CG routine inspections with different Sources. Some inspectors document the Source as “Routines.” Other inspectors will choose “Complaint” as the Source because a complaint is the reason they went to the R­2 in the first place. We were told that inspectors who document their CG routine inspection with “Routines” as the Source do so because only then can the complaint investigation and the CG routine inspection both be ​ ​ counted in CTS. In this instance, CTS can capture both the complaint investigation and the CG routine inspection because they are documented under separate complaint numbers. Also, by choosing “Routines” as the Source, the CG routine inspection will be counted as a routine inspection, not a complaint inspection. We were told that inspectors who use “Complaint” as the Source for their CG routine inspections, will use the same inspection number to report their complaint and CG routine inspection. Under this scenario, the CG routine inspection­­including, the date description—will be buried in the “description” section of the inspection report that primarily documents the complaint investigation. We were told that some district inspectors do not open a new complaint number to document the CG routine inspection because creating a new complaint form takes additional time and they are “too busy.” Fire Safety Inspections in San Francisco 16 When a violation in the common areas is discovered from the CG routine inspection, inspectors will document this differently from one another. If there are violations arising out of both the complaint and CG routine inspection, some inspectors told us that they will open a new ​ ​​ complaint number for the violation in the common areas and change the Source from “Complaint” to “Routines” to document that violation. However, one inspector told us that he will only open a new complaint number and change the Source from “Complaint” to “Routines” if the violations arising from the complaint and those discovered in the common areas have a different anticipated time for compliance. Otherwise, if there are just a few violations in the common area, then he will include them on the same NOV as those arising from the complaint and under the same complaint number. Yet another inspector said that he will not open a new complaint (and, will not change the Source from “Complaint” to “Routines”) to document the CG routine inspection even if there are violations in the common areas. Instead, he will issue a separate NOV for the different violations but will document the the CG routine inspection NOV under the “Description” section of the complaint inspection report.

HIS Does Not Measure The Effectiveness Of Its Code Enforcement Process According to management gurus, W. Edward Deming and Peter Drucker, “you can’t manage what you don’t measure.” Success must be defined and tracked in order to determine whether an organization is successful. For HIS, the definition of success can be defined in two parts. The first part is mandated by Code­­that HIS conduct routine inspections at least every five years. The second part can include ensuring that all violations are corrected within a “reasonable time.” However, HIS does not track when all of the R­2s in our City last had a routine inspection nor does HIS track how long violations take to be corrected. Instead, HIS manages its code enforcement efforts with the goal of moving open violations towards correction without defining what successful correction of violations means. Based on our interviews, we learned that senior inspectors monitor inspectors’ code enforcement efforts by spot checking open cases in CTS and by utilizing a review process every four to six months. We were told that two or three times a year HIS inspectors and senior inspectors are given two management reports, “Complaints Received” and “First NOV Sent,” to review. Both reports are standard reports that can be generated by CTS by senior inspectors or management without the help of DBI MIS. Complaints Received. The first report, “Complaints Received,” tracks complaints that have ​ been received, are still open and for which an NOV has not been issued. The report lists the complaint number, the date the complaint was received, the R­2 address, the date of last activity and a comments section. Each inspector gets their own report with their open complaints listed. Fire Safety Inspections in San Francisco 31 This report provides information on complaints that are open­­it does not include any information on routine inspections. Recently, HIS distributed this report to its 14 district inspectors for open complaints received in 2015. We reviewed a copy of this report. There are 281 complaints received in 2015 that remained open (without an NOV) as of early May 2016. First NOV Sent. The second report, “First NOV Sent,” tracks cases that have had an NOV ​ issued, but, have not been scheduled for a Director’s Hearing. The report is distributed to routine and district inspectors. It includes open NOVs arising from complaints and routine inspections. This report lists the complaint number, the date the first NOV was issued, the R­2 address, the date of last activity and a comments section. Each inspector gets individualized reports with their cases on it. The most recent copy of this report was distributed to inspectors in early May 2016 for open NOVs that were issued in 2015. We reviewed a copy of this report. There were 311 NOVs issued in 2015 that still have not been referred to a Director’s Hearing (“DH”) as of May 2016. ​​ ​​ This means that after at least five months (end of 2015 to May 2016), 311 open NOVs had not been referred to a DH. Although not all of the 311 open NOVs may be appropriate to refer to a DH, this is a significant number of NOVs that are still uncorrected after at least five months. Qualitative review. Once inspectors receive the two reports, they have six to eight weeks to ​​ review their cases. Inspectors provide additional updates on their cases in the comments section of the reports. Afterwards, inspectors meet with their senior inspectors to go over the two reports one case at a time. We were told that the reports are designed to identify open cases and open NOVs that have stalled and encourage inspectors to move the cases along. By the time they meet with their senior inspectors, inspectors may have taken additional actions on many of the open cases (moved them along) on their lists. Consequently, only the more challenging cases are discussed at length with their senior inspector. For these cases, the senior inspectors will help their inspectors determine the next course of action.

District Inspectors Do Not Always Conduct Complaint­Generated Routine Inspections Even though district inspectors are already at the R­2 investigating a complaint, we were told that some are “too busy” with their complaint work to find the time to actually conduct a CG routine inspection or “research” an R­2 before going out there. Based on our interviews, there seems to be an acknowledgement within HIS that district inspectors sometimes are “too busy” and that being “too busy” is an acceptable reason for not conducting a CG routine inspection. This is problematic because CG routine inspections are a convenient way to conduct routine inspections. SOP does not explicitly require that CG routine inspections be conducted. The Standard ​ Operating Procedure (“SOP”) is used to train all HIS staff (including inspectors) and includes detailed procedures for conducting inspections and code enforcement. Although we were told that district inspectors should be conducting CG routine inspections of the common areas of an ​ ​ R­2 when investigating a complaint, the SOP does not explicitly require it. Instead, the SOP is very vague and only requires inspectors to “schedule site inspection appointment.” It does not mention what kind of “site inspection” should be performed. The SOP also does not address whether “the site inspection” should include common areas even if they are not part of the complaint.29 Inspectors should “research” properties before an inspection. Additionally, we were told ​ that district inspectors should be “researching” R­2s in CTS before going out to investigate a complaint so that district inspectors will know when an R­2 is due for a CG routine inspection. What exactly should be “researched,” however, is subject to different interpretations. When researching an R­2, some district inspectors only look at the history of complaints on an R­2 ​ ​ while others also research when the last routine inspection was performed on an R­2. All ​​ ​ 29 Housing Inspection Services Policies and Procedures Manual, , Item 4. Fire Safety Inspections in San Francisco 17 inspectors have smart phones with internet access so they can be used to access CTS records on the DBI website. 5. “No Shows” Waste Inspectors’ Time A “no show” is when the property owner fails to appear for a scheduled routine inspection appointment. Unless the inspector finds another way to get into the R­2 to conduct the routine inspection, a “no show” will mean that the inspector wasted precious time going out to the R­2 and that at least one additional visit to the R­2 will be necessary. Furthermore, the need to still perform a routine inspection may get lost and the R­2 in question may not have a routine inspection within the mandated five year time period. “No shows” are not tracked for follow­up in CTS. Currently, CTS is not being used to track ​ ​ “no shows” for HIS. Although CTS can track no shows when inspectors choose either “No Entry” or “Unable to Enter” as the “Abatement Type,” we were told that inspectors were instructed not to use either of these. We were told this is because by using “No Entry” or “Unable to Enter” the inspector’s attempt to make an inspection would not be counted as an inspection. Instead, inspectors were instructed to use another “Abatement Type” to reflect the type of inspection they made. For example, “Inspection of Premises Made” or “Reinspection 1.” As a result, inspectors document “no shows” as part of a narrative in the description section of the complaint form. However, once “no shows” are buried in the description section, there is no way to run a report on “no shows” or flag them for follow­up in CTS. Inspectors have shared that, as a result, they may “lose track” of these “no shows” as their workload requires them to direct their efforts elsewhere. Measuring the extent of “no shows.” As part of its Focused Code Enforcement, HIS started to ​ track “no shows” (along with other results of its routine inspections) manually on an Excel spreadsheet. We have included this data in Table I­3 but only for the three districts for which HIS has compiled this information. Table I­3 shows the number of routine inspection appointment letters that were sent to property owners during 2015, the number of R­2s for which inspectors were not able to conduct routine inspections because they were unable to enter due to “no shows” and the percentage of R­2s that had “no shows” as a percentage of the total inspection appointment letters sent. FOCUSED CODE ENFORCEMENT “NO SHOWS” 2015 District Number of inspection Number of R­2s unable Percent R­2s unable appointment letters sent to enter (“no shows”) to enter (“no shows”) Mission St. Corridor 128 20 16% Chinatown 167 28 17% Mission* 259 38 15% *Does not include R­2s on Mission Street Table I­3 Fire Safety Inspections in San Francisco 18 Clearly, the percentage of R­2s that inspectors were unable to enter (“no shows”) during their Focused Code Enforcement is significant for all three areas for which this information was captured. It’s difficult to extrapolate this data and assume similar percentages exist in other districts in the City. However, it makes one wonder if these are not the only areas with a significant “no show” problem. If this happens time after time, (as we learned it sometimes does) these “no shows” and their impact on the inspection backlog really start to add up. We were told that “no shows” occur mainly when inspection packets go to an incorrect address. From our own observation, we concluded that they also happen because the inspection packets can be difficult to understand. Inspection packets that are sent to property owners sometimes go to an incorrect address. Support staff schedule a group of routine inspections for R­2s that have not been inspected within the last five years. As part of the scheduling process, HIS support staff send out inspection packets to the property owners of record. This information comes from the Tax Assessor database to which HIS has access. Sometimes the inspection packets go to the wrong address and are returned to HIS. (We’ve been told that Tax Assessor records may be outdated by as much as 18 months but we have not independently verified this.) Inspection packets are only sent to property owners in English. The inspection packet ​ includes a cover letter stating that a “periodic health and safety inspection” will be conducted in the common areas of their building, the authority for performing the inspection, and the scheduled time and date for the inspection. The scheduled inspection date gives the property owner two weeks notice and may be changed by contacting the listed inspector by email, phone, or in person. The inspection packet also includes the following: (1) Property Owner Maintenance Checklist (which is the same as the Inspection Field Checklist); (2) Notice of New Housing Law regarding wood fixed utility ladders; (3) Ordinance 255­08; (4) handout on New Ban on Wooden Fixed­Utility Ladders; (5) Notice Requiring Compliance of San Francisco Housing Code Section 604; (6) Compliance Affidavit; and (7) Affidavit­Self Certification for Carbon Monoxide and Smoke Alarms. Although the Property Owner Maintenance Checklist is available on the DBI website in Chinese and Spanish,30 the inspection packet is only provided in English to property owners. The inspection packet is difficult to understand. Although the cover letter contains vital ​ information, much of the information is buried in the body of the letter. Also nowhere in the letter or packet is it explained what the inspector will be inspecting. Instead the Property Owner Maintenance Checklist is merely included with no explanation for its purpose. Furthermore, the appendage and carbon monoxide/smoke alarm affidavits are included in the packet without instructions on what needs to be done with them­­they should be filled out and returned to HIS. Lastly, including the Notices, Ordinances and informational flyers is confusing because they are not tied in with the rest of the inspection packet. (See Appendix, Exhibit 11.) 30 Property Owner Maintenance Checklist http://sfdbi.org/sites/default/files/Checklist%20English.pdf ​ Fire Safety Inspections in San Francisco 19 “No shows” frequently are not followed up on. “No shows” are not uncommon and may occur ​ ​​ scheduled routine after scheduled routine on the same property. On occasion, an inspector may be able to get current contact information (including a phone number) for the property owner (or property management company) from a tenant at the R­2. In these instances, the inspector will try to reschedule the routine inspection using the updated information. We were told that unless a property owner calls to reschedule, it is much more common that inspectors and support staff will not follow up on the “no shows” because they do not have the time necessary for researching the property further.

HIS Lacks More Effective Code Enforcement Tools Although HIS has a well established code enforcement system that effectuates timely abatement in many cases, there is a common belief among HIS inspectors that their code enforcement tools often may not be effective enough. When inspectors lack effective tools to motivate the reluctant property owner to abate violations more quickly, the enforcement period may be unnecessarily extended. This means inspectors have to work harder by conducting many more reinspections and other tasks in hopes of achieving abatement. There is also the looming possibility that the violation will still not be corrected. Furthermore, extending the time for abatement exposes our housing stock, its tenants and neighbors to unnecessary risks that should have been corrected sooner. HIS’ most effective tools for incenting abatement include referring cases to a DH or the City Attorney and the Special Assessment Lien program. (See Appendix, Exhibit 12, Explanation of HIS Code Enforcement.) Their effectiveness is largely due to the high costs and negative publicity associated with these programs. City Attorney cases apply further financial incentive due to the possibility of multiple civil penalties, punitive penalties and attorney's fees being awarded. Even so, these tools may not be effective in every case. SOP, , Item 9(c). San Francisco Building Code, Chapter 1A and SOP ( , item #12(c)). ​​ ​​ Fire Safety Inspections in San Francisco 34 We were told that the most stringent (and effective) determination coming out of a Director’s Hearing­­the Order of Abatement­­ may have lost much of its deterrent potential. In the past, property owners did not want their property title to be clouded with an Order of Abatement as it would make borrowing money and selling the property more difficult. Nowadays, with the change in the financial markets, we were told there aren’t the same obstacles connected with a clouded title as there have been in the past. Additionally, unless a property owner intends to borrow money or sell the property, they could choose to defer abatement for many years. Cases that are referred to the City Attorney almost always show results­­the violations are almost always cured. For the period 2010 through 2015, DBI referred 59 cases to the City Attorney. Nine of these cases, referred in 2010, were for buildings owned by one firm, Blanding (doing business as Bayview Property Managers). The City Attorney obtained an injunction requiring Blanding address nuisance conditions at all 30 buildings they own and/or manage and also imposed a civil penalty of $800,000. Ten cases dealt with vacant or dilapidated buildings, several occupied by elderly owners who are no longer able to care for their homes. Fourteen violations referred to the City Attorney (“CA”) were for work without permit, including adding illegal units. In one case, a parking garage owner turned one floor into a hostel. Eighteen cases dealt with multiple violations, including fire safety and structural damage. Almost all were abated within a few months of the City Attorney’s involvement, although one case took over two years. It required a restraining order on the owner, and the bank holding the mortgage placing the property in receivership. However, the CA pursuing litigation can be a very expensive route, and therefore, it is reserved for the most egregious cases. Lastly, the Special Assessment Lien program is also very effective but is only available once a year. Franchise Tax Board. Years ago the California Franchise Tax Board (“FTB”) had a program ​ that we were told was very effective in getting property owners to correct violations. Under this program, once a violation had gone uncorrected for 180 days after the initial NOV compliance period had elapsed (usually thirty days), the inspector could refer the case to the FTB. After the inspector received approval to refer a case to the FTB from their senior, they would prepare a Notice of Non­Compliance. The Notice of Non­Compliance would be recorded and sent to the property owner and the FTB. As a result, when the property owner filed their California tax return and attempted to take deductions for expenses incurred in connection with their rental property (the R­2 with the NOV), the FTB would disallow these deductions until the NOVs were abated. In response, property owners would correct the previously unabated violations. Thereafter, a Notice of Compliance would be issued and sent to the property owner, the Recorder’s office and the FTB. Administrative penalties. Currently, HIS cannot impose civil penalties on property owners for ​ unabated violations because the current administrative hearing HIS uses (the Director’s Hearing) does not comport with due process requirements that are necessary for civil penalties to be awarded. There’s a belief within HIS that being able to seek administrative civil penalties would create a significant financial incentive for property owners to abate violations. Since a case can be referred to a DH as early as sixty days after an NOV is issued, the threat of administrative penalties being awarded may encourage property owners to correct violations more quickly. We Fire Safety Inspections in San Francisco 35 were told that, in order for administrative civil penalties to be added to the administrative hearing, at a minimum, the following would have to change: (i) the hearing officer would have to come from outside DBI; (ii) testimony may have to be given under the penalty of perjury; and, (iii) the notice of decision would have to have an appeal process that first went to a board, comprised of members other than the BIC. HIS could seek administrative civil penalties when there were repeat offenders, serious deferred maintenance, numerous NOVs, or a vulnerable population was being affected, along with many more cases. FINDINGS ​ F.I.29. HIS does not measure how long NOVs take to be abated. Without tracking how long it ​ takes for NOVs to be abated, HIS cannot determine whether it’s code enforcement process is effective for correcting all violations in a timely manner. F.I.30. For 2013­2015, approximately twenty percent of NOVs took more than one year to correct. F.I.31. HIS does not have a standard against which inspectors’ grant of additional time can be ​ measured. F.I.32. When inspectors grant additional time for property owners to correct an abatement, there is no written documentation (other than on an NOV) provided to the property owner that states when the next reinspection will occur or explains that violations must be abated by then. By not communicating this in writing, property owners make think that they can negotiate with the inspectors more easily. Also, some property owners may not understand what they are being told due to language differences or other reasons. F.I.33. Although bi­monthly staff meetings are scheduled, they are regularly cancelled because inspectors are “too busy.” Without a management culture that supports having scheduled times to discuss inspectors work, it will be difficult for HIS to optimize its code enforcement process for success. F.I.34. Based on our investigation, we concluded that HIS does not have an adequate definition for success. F.I.35. Some inspectors take too long to refer open NOVs to a DH. But, HIS does not measure ​ ​ how long it takes an open NOV to reach a Director’s Hearing. F.I.36. Inspectors take too long to refer open NOVs to a DH because the standard for referring unabated violations to a Director’s Hearing is vague and leaves too much room for interpretation. F.I.37. Not all inspectors proactively brief their seniors after three reinspections with no progress. Fire Safety Inspections in San Francisco 36 F.I.38. Inspectors take too long to refer open NOVs to a DH because preparing a case for referral to a Director’s Hearing is more labor intensive than it should be. F.I.39. HIS lacks more effective code enforcement tools.

CTS Is Outdated CTS is a legacy system that lacks capabilities that are commonplace in today’s workplace. Without these capabilities, inspectors and support staff must spend much more time doing tasks ​ that would take less time with a more robust computer system. This loss of efficiency contributes to the difficulty of not being able to conduct routine inspections on all R­2s in San Francisco within the mandated five year timeframe. Also, it results in violations that take longer to correct because inspectors do not have the time available or tools necessary to monitor their cases sufficiently. We find these capabilities missing: ● CTS cannot be accessed from the field. Inspectors cannot input data to CTS from ​ outside the office. As a result, inspectors must document inspections twice. Inspectors ​ document routine inspections at the inspection site by taking handwritten notes­­typically on their Inspector Field Checklist. When the inspector arrives back at the office he will type up the written notes into CTS and upload any photos taken at the inspection site into the network “P” drive. The “P” drive is a separate drive that is not connected to CTS nor can it be accessed outside the office. ● Affidavits are not available online. Currently, the appendage and carbon ​ monoxide/smoke alarm affidavits are not available on the DBI website. The appendage affidavit must be completed by a licensed or certified professional stating that all appendages to an R­2 are structurally safe. The affidavit is due every five years. The carbon monoxide/smoke alarm affidavit states that carbon monoxide and smoke alarms have been installed in compliance with the code and must be signed by the property owner. Also, the affidavits cannot be completed by hand, scanned and sent digitally to HIS. ● Notices of Violation (“NOVs”) cannot be printed in the field. Inspectors cannot print ​ and post the NOV while at the R­2 because CTS cannot be accessed remotely. Also, inspectors do not have portable printers. Therefore, the inspector must return to the R­2 to post the NOV on the building. ● CTS is not integrated with computer systems within DBI or other City departments. HIS cannot share data across departments­­most importantly within DBI, Department of ​ Health (“DPH"), Department of Public Works (“DPW”) and SFFD­­so that it can ​ ​ ​ coordinate its inspection and code enforcement efforts and reduce redundancies. Also, HIS cannot know when permits have been filed for and approved and the scope of Fire Safety Inspections in San Francisco 20 permits so that inspectors can have insight into property owners’ efforts to correct violations. In addition, HIS cannot create its R­2 routine inspection lists without having to ask DBI MIS for this information. ● CTS cannot track and report on important attributes. CTS’s ability to track and ​ report on important attributes, such as type of violations and building characteristic is limited. Currently, inspectors document the type of violation from a drop­down menu that offers 83 options in CTS. We were told that, oftentimes, NOVs list multiple violations of different types and that CTS cannot track individual violations listed on an NOV. CTS only has the capability to track the overall NOV. As a result, when an NOV lists multiple violations of different types, inspectors will document the NOV as “General Maintenance” for the type of violation, rather than the specific type of violations listed on the NOV. Similarly, from our review of CTS input screens, we learned that inspectors can document the overall condition of an R­2 by ranking it on a scale from one to five. However, CTS cannot track factors that are useful in determining which R­2s have higher fire risks. We were told that some of the factors that should be to considered when looking at an R­2’s fire risk include: (1) the age of the building; (2) the materials used to construct the building; (3) the overall condition of the building; (4) whether the building has a fireblock; (5) whether the building is particularly densely populated due to illegal tenants; (6) whether the building is of mixed use (residential and commercial); and, (7) whether the building is located on a major thoroughfare. We concluded, from our interviews and review of CTS input screens, that CTS has not been updated or revised to better meet HIS’ needs because DBI believed Accela, which was initially scheduled to be implemented in 2013, would resolve any and all issues HIS had with CTS. Accela. In October 2011, the City entered into a $4.5 million contract with two information ​ technology companies, Accela and 21 Tech, to build and implement an integrated computer system (“Accela”) that would replace the Planning Department (“Planning”) and DBI’s legacy systems. In essence, Accela was intended to streamline the permit process by enabling Planning and DBI to seamlessly share data across departments and provide online access so that the permit process could be conducted online with transparency. After the initial roll­out to Planning and DBI, Accela would then replace legacy systems in other City departments, including the SFFD, Department of Public Works and Department of Public Health, among others. As part of this process, CTS was going to be replaced by Accela which promised to offer (1) integrated data sharing across HIS, other DBI divisions and Planning; (2) more tailored and automated report functionality; (3) more extensive data point tracking; (4) online capabilities; and, (5) automation of manual processes. Accela was first scheduled to go live for Planning and DBI in late 2013. However, this launch was postponed. From late 2013 to late 2015, change orders for Accela were numerous and were estimated to increase the cost of the Accela roll­out by close to $4 million (which turned out to be accurate based on Gartner’s, a third party vendor, finding that change orders raised the cost to Fire Safety Inspections in San Francisco 21 $8,174,297).31 Several other launch dates were scheduled and postponed during this time. Consequently, in late 2015, DBI put Accela on hold. Third party outside vendors, Cosmo Cloud and Gartner, were contracted to perform requirement scoping and determine how much more time, money and work would be needed for Accela to be successfully implemented. They also were to evaluate whether implementing Accela was viable and the optimal choice. The reasons why Accela could not be implemented within its contract price on its original launch date are beyond the scope of this report because we have not specifically verified them. For those interested, reasons do appear in the Gartner report issued on June 9, 2016.32 According to the report, implementing Accela will require negotiating a contract amendment, addressing gaps in the off­the­shelf system, and strong support from DBI throughout the development process, with go­live estimated to occur between August and October 2017. FINDINGS F.I.1. Housing Inspection Services (“HIS”) does not know which R­2s have not been inspected within the last five years because the Complaint Tracking System (“CTS”) cannot generate a list of R­2s with an accurate last routine inspection date for each. F.I.2. The spreadsheet used by HIS to track key inspection statistics has not been updated to include all rounds of Focused Code Enforcement completed to date. F.I.3. Because “Routine Inspections” that are reported to the Building Inspection Commission on a monthly basis include the number of initial routine inspections and reinspections ​ ​ that have been conducted, this performance measure is misleading. The total number of initial routine inspections that have been conducted is the correct statistic for determining how many R­2s have had the Code mandated routine inspection at least every five years. F.I.4. HIS cannot get an accurate list of R­2s in the City without the help of DBI Management ​ Information Systems (“DBI MIS”) because HIS does not have access to the DBI ​ database that stores this information. F.I.5. DBI MIS doesn’t always generate the initial list of R­2s, including the property’s address and property owner’s contact information, for HIS. F.I.6. The final list of R­2s for routine inspections is created manually because inspectors and/or support staff must look up the date of the last routine inspection for each R­2. When inspectors do this, it takes them away from conducting inspections. ​ 31 Gartner Report, http://sfdbi.org/sites/default/files/BIC%20Meeting%2006­15­16%20Agenda%20%2314.pdf ​ 32 Gartner Report, http://sfdbi.org/sites/default/files/BIC%20Meeting%2006­15­16%20Agenda%20%2314.pdf ​ Fire Safety Inspections in San Francisco 22 F.I.7. Although the routine inspection backlog that existed in the Mission, Chinatown and Tenderloin Districts has been reduced through Focused Code Enforcement, a routine inspection backlog still exists in these areas. F.I.8. Inspectors do not choose the same “Source” and “Abatement Type” when documenting routine inspections. Unless all the possible ways to document a routine inspection are known and CTS report parameters are chosen to capture all the possible alternatives, some routine inspections will not be captured by a report purported to list all routine inspections. F.I.9. Since CTS does not have “Complaint Generated Routine” as an option for documenting the “Source” for CG routine inspections, CTS cannot separately track and report on complaint­generated routine inspections (“CG routine inspections”). F.I.10. Inspectors do not choose the same “Source” when documenting CG routine inspections. When inspectors choose “Complaint” as the Source, the CG routine inspection will not be counted as a routine inspection in CTS, and HIS will not have an accurate last routine inspection date for those R­2s. F.I.11. District inspectors do not always conduct a CG routine inspection while they are investigating a complaint at an R­2 even when the R­2 has not had a routine inspection for five years because they are “too busy.” HIS accepts inspectors being “too busy” as an excuse for not conducting a complaint­generated routine inspection. F.I.12. HIS’ Standard Operating Procedure (“SOP”) does not explicitly require inspectors to conduct a CG routine inspection while they are investigating a complaint at an R­2 when the R­2 has not had a routine inspection within the last five years. F.I.13. District inspectors do not always know when an R­2, at which they are investigating a complaint, is due for a complaint­generated routine inspection because there is no clear requirement to “research” the last routine inspection date before investigating a complaint. F.I.14. Inspectors cannot always get into an R­2 to perform a scheduled routine inspection because of “no shows.” Since CTS cannot track “no shows,” inspectors sometimes lose track of the fact that a routine inspection still needs to be conducted on the R­2s that have a “no show.” F.I.15. HIS has started to manually track “no shows” on an Excel spreadsheet that tracks results of their Focused Code Enforcement. However, this spreadsheet has not been completed for all routine inspections conducted under Focused Code Enforcement. F.I.16. There was a significant number of inspection “no shows” in the Chinatown (17%) and Mission (15%) Districts and in the Mission Street Corridor (16%). Oftentimes “no shows” are not followed up on because staff is “too busy” to research the property owner’s correct address or phone number. Fire Safety Inspections in San Francisco 23 F.I.17. Inspection packets that are sent to property owners sometimes go to an incorrect address because data provided by the Tax Assessor’s Office does not have up­to­date contact information for the property owner. F.I.18. Inspection packets are sent to property owners only in English. F.I.19. The inspection packet cover letter is confusing and buries vital information in the text. F.I.20. The Property Owner Maintenance Checklist included in the inspection packet is not explained as being the list of items that will be inspected. F.I.21. Instructions on what the property owner needs to do with the appendage and carbon monoxide/smoke alarm affidavits included in the inspection are not included on the affidavits or elsewhere in the inspection packet. F.I.22. Including notices, ordinances and information flyers in the inspection packet without explaining their purpose is confusing. F.I.23. Inspection documentation is done twice (first in the field and again into CTS when the inspector returns to the office) because there is no online access to CTS. F.I.24. Photos cannot be uploaded into CTS because CTS does not have this functionality. Instead, they are stored on the network “P” drive which is not connected to CTS. F.I.25. Affidavits are not available online. F.I.26. Inspectors are not able to print NOVs in the field. Therefore, they must return to the property a second time to post the NOV on the R­2. This is a waste of time and resources. F.I.27. CTS is not integrated with computer systems within DBI or other City departments. F.I.28. CTS cannot track and report on important attributes, such as types of violations and high fire risk building characteristics.