Identification of Liability-Related Impediments to Sharing §409 Safety Data among Transportation Agencies and a Synthesis of Best Practices (2006)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Research Results Digest 306 October 2006 C O N T E N T S 1. Introduction, 1 2. Understanding the Problem, 2 3. Achieving Successful Data Sharing, 5 4. Managing the Risk of Liability, 6 5. Conclusions, 10 References, 11 Endnotes, 11 Acknowledgments, 11 Appendix A, 12 Appendixes B through E, 14 1. INTRODUCTION This digest identifies liability risks as- sociated with sharing safety data among transportation agencies pursuant to Sec- tion 409 of Title 23, U.S.C.; identifies best practices; reviews the Pierce County, Washington v. Guillen decision and its potential impact on managing state lia- bility risk; and describes strategies for overcoming the impediments to data shar- ing, specifically those related to liability. Transportation lawyers and risk manage- ment practitioners for the states, metro- politan planning organizations (MPOs), and transportation planning organizations may find this digest useful. Background Successfully implementing safety im- provements and integrating safety into the transportation planning process requires that crash data be made available to local communities, MPOs, and a range of both traditional and nontraditional safety part- ners. Complete, accurate, and timely data are necessary for correctly identifying the size, location, and characteristics of safety problems; examining alternative solution strategies and tradeoffs among them; and evaluating the effectiveness of solutions and treatments. Since the early 1960s, federal programs have encouraged states to collect and use crash data to develop methods for identify- ing safety problems and prioritizing funding for safety improvements. In 1987, Congress enacted 23 U.S.C. §4091, which states that data compiled and collected, inter alia, pur- suant to Title 23, U.S.C. §1521 cannot be used as evidence against a state in litiga- tion, including discovery. The constitu- tionality of §409 was upheld by the U.S. Supreme Court in Pierce County, Wash- ington v. Guillen 537 U.S. 129; 123 S.Ct. 720 (2003), as summarized in Appendix A. Despite Guillen, a number of states con- tinue to withhold or restrict the sharing of data. Approach The project team conducted a series of interviews with six states identified by the NCHRP panel. Within each of these states, interviews were conducted with a plan- ning representative, the person responsible for managing safety data, and/or the lawyer IDENTIFICATION OF LIABILITY-RELATED IMPEDIMENTS TO SHARING §409 SAFETY DATA AMONG TRANSPORTATION AGENCIES AND A SYNTHESIS OF BEST PRACTICES This digest presents the results of NCHRP Project 8-54, “Identification of Liability-Related Impediments to Sharing §409 Safety Data Among Transportation Agencies and a Synthesis of Best Practices.” The research project was conducted by Cambridge Systematics, Inc., with support from the American Automobile Association. Subject Areas: IA Planning and Administration, IC Transportation Law Responsible Senior Program Officer: Gwen Chisholm-Smith NATIONAL COOPERATIVE HIGHWAY RESEARCH PROGRAM

responsible for data-related liability concerns. Given the objective of determining best practices for over- coming liability-related impediments to the sharing of crash data, the interviews focused on determining the states’ current data-sharing practices; the reasons for limitations, if any, in the sharing of crash data; the degree to which any limitations are related to liabil- ity concerns; and the practices employed in manag- ing the liability-related risks in sharing safety data. Survey questions also were made available to an additional 19 states through the use of an Internet- based electronic survey. Consistent with the informa- tion obtained in the state interviews, these responses indicated that (a) it is the policy of some states to share crash data with their planning partners while others are more reluctant; (b) a few respondents either were not familiar with the Guillen decision or had not analyzed the decision in any detail; and (c) states are using a variety of means to manage the risk of sharing data. Report Structure Section 2 provides a brief legislative history of the use of crash data in safety planning, how safety has emerged as a key planning factor, and a descrip- tion of federal safety funding programs. Findings of interviews and surveys are summarized in Section 3. Section 4 provides methods through which states are managing the risk of sharing safety data. Section 5 provides overall conclusions regarding the further examination of the protection provided under §409 and resources for managing the risks associated with sharing safety data. An analysis of the U.S. Supreme Court’s Pierce County, Washington v. Guillen deci- sion is provided in Appendix A. The interview and survey questions are included as Appendix B. Copies of representative pertinent state documents gathered as a part of this research are contained in Appendix C, with an example of a state Freedom of Information Act provided as Appendix D.2 A paper on the shar- ing of safety data prepared by the Transport Research Centre of the Netherlands Ministry of Transport is provided as Appendix E. Appendices B through E, not published herein, are available on the TRB web- site as NCHRP Web-Only Document 89. 2. UNDERSTANDING THE PROBLEM This section provides a summary of the legisla- tive history of federal safety funding programs; de- scribes the significance of collecting and analyzing safety data and how it can be used by multiple plan- ning partners; and examines the issues states face when trying to share these data. Brief Legislative History of Federal Safety Funding Programs Early federal roadway funding was devoted pri- marily to new capital construction, with safety incor- porated both in design standards and at the time of any upgrade or rebuild. Beginning in the mid-1960s, the government provided funds to state and local gov- ernments to explicitly make safety improvements on existing roads in an effort to help reduce the number and severity of motor vehicle crashes. The Highway Safety Act of 1973, for the first time, provided cate- gorical safety funds for infrastructure-related high- way safety improvements. Implementing the new federal safety program required the states to collect safety-related data that could be used to identify safety problems and priori- tize funding for safety improvements. The data were collected in police accident reports and included crash location data, roadway features, collision de- tails, injury information, and other details. About the same time that new safety programs were established, state court decisions and legisla- tion began limiting, and even waiving, sovereign immunity from tort lawsuits. States that did not have broad sovereign immunity realized the inherent con- flict between the data collection elements of the re- quired safety programs and potential tort liability. Without immunity, there was a potential liability for transportation agencies on claims against the state for improper facility design and failure to upgrade or improve facilities that appeared to be unsafe based on an analysis of crash data. Some states effectively limited their transportation liability by adopting laws and rules exempting certain government decisions; however, others did not, which exposed them to broad liability claims. The Federal highway safety laws created liability problems in states with broad transportation lia- bility. The extensive collection of safety-related data on existing highways allowed plaintiffs to argue that states had notice of ‘defects’ on exist- ing roads. Plaintiffs then argued that notice of the defect created a legal obligation to ‘repair’ or ‘improve’ the ‘defective’ road. The safety data that is most problematic for highway agencies was the database of collision reports coded 2

by location, accident characteristics, and high- way features. This database made it very easy for plaintiffs to argue that highway agencies had notice of roadway conditions which had caused prior deaths and severe injuries, but had not been ‘repaired.’ (Michael Tardif, Analysis of Pierce County v. Guillen and Its Effect on Sharing of Transportation Data, Washington State Attorney General’s Office.) In 1987, Congress enacted 23 U.S.C. §409, and subsequently amended it, to mitigate the risk asso- ciated with sharing safety data. Title 23 was sub- sequently amended by the Intermodal Surface Trans- portation Efficiency Act of 1991 (ISTEA); the National Highway System Designation Act of 1995; and the Transportation Equity Act for the 21st Cen- tury (TEA-21) in 1998. Until the 2005 surface trans- portation reauthorization, §409 read: Notwithstanding any other provision of law, reports, surveys, schedules, lists, or data com- piled or collected for the purpose of identifying, evaluating, or planning the safety enhancement of potential accident sites, hazardous roadway conditions, or railway-highway crossings, pur- suant to sections 130, 144, and 152 of this title or for the purpose of developing any highway safety construction improvement project which may be implemented utilizing Federal-aid highway funds shall not be subject to discovery or admitted into evidence in a Federal or State court proceeding or considered for other purposes in any action for damages arising from any occurrence at a loca- tion mentioned or addressed in such reports, sur- veys, schedules, lists, or data. This statute was intended to protect from disclo- sure, at trial or in pretrial discovery, data compiled and collected by state DOTs pursuant to Title 23 U.S.C. §152. In other words, data protected by §409 cannot be used as evidence against the state to prove, for example, claims that the state had prior knowledge of a dangerous road and did not improve it, or that the state did not properly prioritize its dangerous roads for improvement, when the state is being sued for alleged liability in connection with a crash or other event. Transportation Equity Act for the 21st Century The Transportation Equity Act for the 21st Cen- tury (TEA-21) was enacted on June 9, 1998, and authorized the federal surface transportation programs for highways, highway safety, and transit for fiscal years 1998 through 2003. The Act focused much more significant attention on safety than previous transportation funding acts. TEA-21 designated “the safety and security of the transportation system for motorized and nonmotorized users” as one of the seven key areas to be considered in the overall trans- portation planning process, at both the metropolitan and statewide levels. This designation helped raise awareness of safety, encouraged collaboration among metropolitan and statewide transportation planning officials, and enforced the need for collection and sharing of crash data. Safe, Accountable, Flexible, Efficient Transportation Equity Act: A Legacy for Users On August 10, 2005, the most recent trans- portation reauthorization bill, known as the Safe, Accountable, Flexible, and Efficient Transportation Equity Act: A Legacy for Users (SAFETEA-LU) was signed into law. This legislation further elevated the importance of safety within the transportation planning process. The Act established the Highway Safety Improvement Program, a new core funding program, under which states are required to develop a Strategic Highway Safety Plan (SHSP). The SHSP is to be data-driven and developed through a com- prehensive and collaborative process. Under §148 of Title 23, U.S.C., members of the SHSP planning process must include the following: • Governor’s Highway Safety Office; • Regional transportation planning organizations and MPOs; • Major modes of transportation; • State and local traffic enforcement officials; • State persons responsible for administering the federal rail-grade crossing program; • Representatives conducting Operation Life- saver; • State Motor Carrier Safety Assistance Program (MCSAP) administrators; • State motor vehicle administrators; and • Major state and local stakeholders. States must meet several requirements governing the obligation of funds under §148. One requirement is to submit to the Secretary an annual report that: (i) describes, in a clearly understandable fashion, not less than five percent of locations determined 3

by the State, using criteria established in accor- dance with paragraph (2)(B)(ii), as exhibiting the most severe safety needs; and (ii) contains an as- sessment of—(I) potential remedies to hazardous locations identified; (II) estimated costs associ- ated with those remedies; and (III) impediments to implementation other than cost associated with those remedies. Recognition of the importance of data sharing is reflected in the Act and, along with preserving §409, SAFETEA-LU also protects data collected pursuant to the Act’s new data collection provisions in a pro- vision similar and comparable to §409. Subtitle D of §148 states: (4) Discovery and admission into evidence of certain reports, surveys, and information— Notwithstanding any other provision of law, reports, surveys, schedules, lists, or data com- piled or collected for any purpose directly relat- ing to paragraph (1) or subsection (c)(1)(D), or published by the Secretary in accordance with paragraph (3), shall not be subject to discovery or admitted into evidence in a Federal or State court proceeding or considered for other purposes in any action for damages arising from any occur- rence at a location identified or addressed in such reports, surveys, schedules, lists, or other data. Importance of Data Sharing The science of safety demands careful examina- tion of data to inform investment and countermeasure decisions. Historically, safety investment decisions were made based on engineering design standards and experience. Today, safety investment deci- sions are more likely to be data-driven and evaluated for their level of effectiveness for reaching stated goals and objectives, using crash data for a specific lo- cation, corridor, or region. Through an open planning process, it can be expected that better problem identi- fication leads to a greater likelihood that resources will be used to address the most serious safety prob- lems. Safety, therefore, is more likely to receive equal and explicit consideration with other transportation planning priorities. Also, data analysis helps to ensure that the most effective and efficient treatments and countermeasures will be deployed because data allow quantification of the costs and benefits of alternatives. Current data-sharing practices among states, however, are widely diverse. Some states share some data, but are careful about what data they share and with whom. The challenge to mitigating the risk of sharing data is further complicated by the various audiences who request or need access to data, such as MPOs, local planning agencies, law enforcement, educators, private consultants, the public, and the media. The source of data requests and the relation- ship with the DOT often determine which means of mitigating risk the DOT will implement. For exam- ple, they may summarize the data before sharing it with the media to ensure accurate reporting. Some states share crash data freely. In one par- ticular state, both state and local agencies need only make a phone call to request data, and the state DOT either prepares a report or provides raw data depend- ing on the user and the request. A few states allow online access to crash data. As described in Section 4, many states that share data freely do so because they have certain protections in place. In one state DOT that makes crash data available to meet the needs of its partners and stakeholders, the planning authority developed a set of regional safety priorities for a specific set of counties as part of that region’s long-range transportation plan update. Par- ticipants in this effort included the local towns and cities, the counties, a local university, state and local police, and health organizations. The effectiveness of using data for transporta- tion—and safety—planning purposes is being pro- moted through several initiatives. For example: • Enactment of SAFETEA-LU established a new core safety infrastructure program, the High- way Safety Improvement Program. Under this program states have until October 1, 2007, to develop a comprehensive, data-driven strate- gic highway safety plan. • The American Association of State Highway and Transportation Officials (AASHTO) has created a strategic highway safety plan and, along with many partners, is encouraging all states to use a collaborative, comprehensive, data-driven process for developing a statewide comprehensive highway safety plan. As with safety conscious planning (SCP) or safety integration, crash data are to be analyzed to identify the key emphasis areas for any state. • An international scan on data issues was re- cently conducted to identify best practices for collecting, managing, analyzing, and sharing crash data. The work of the Transport Research Centre of the Netherlands Ministry of Trans- port is presented in Appendix E as one exam- ple of current international thinking. 4

State Liability Risk American jurisprudence has long recognized a right to redress negligence. Proving liability, in very simple terms, requires the establishment of a duty, the breach of that duty, and damages or injury resulting as a proximate cause of that breach. Factors establish- ing a breach of a public entity’s duty include (1) that which causes the damage must be in the care or cus- tody of the public entity; (2) there was a hazardous condition; (3) the public entity had actual or construc- tive knowledge of the hazardous condition; and (4) it failed to correct the condition within a reasonable pe- riod of time. Based on the facts presented, a jury de- termines whether a duty has indeed been established, whether that duty has been breached, and whether that breach proximately caused damages or injury. America has become an increasingly litigious society over the past few decades. When crashes occur, and especially when people are seriously in- jured or killed, plaintiffs’ attorneys often try to pre- sent facts that suggest state negligence in maintain- ing or improving the road system. If data can be used to show that an accident happened at a “dangerous” intersection, for example, it may then be argued that the state had prior knowledge of the danger, had a duty to address the danger, is negligent for not hav- ing made repairs or improvements, and that negli- gence resulted in damages to the plaintiff. Important questions are: (1) Do states have ade- quate protection through §409 for data collected pursuant to federal statutory provisions; (2) How does the protection provided by §409 relate to a state’s freedom of information or public records act; (3) What are some potential strategies for effectively managing the potential legal risks in sharing crash data so that data-driven safety planning can be en- sured; and (4) Does tort law in a state create prob- lems for achieving the goals of federal safety pro- grams? For example, some states reported that the advent of the federal safety program had the effect of changing state tort law. The availability of federal funding to make safety improvements and the asso- ciated safety data systems used to identify existing roads as “hazardous” or “defective” resulted in courts in these states allowing liability for failure to make capital improvements. 3. ACHIEVING SUCCESSFUL DATA SHARING A data-driven process allows a planning agency to identify needs, examine potential countermeasures, and measure performance of those solutions upon implementation. DOT officials, therefore, have an opportunity to improve planning processes statewide by educating all users (and potential users) on what data are available, how to analyze the data, and how it should and should not be used. This section sum- marizes the administrative or procedural issues and barriers to data sharing as identified by DOT repre- sentatives through interviews and survey responses. In many cases, miscommunication or a lack of com- munication and/or fear of liability were cited as the most significant impediments to data sharing. Timely Access to Data Access to data and the timeliness of data are com- mon complaints of safety planners. In many states, collection and management of crash data are not a DOT responsibility. The data are entered and main- tained by another state agency, such as an office of safety or a law enforcement agency. In some cases, crash/collision data are entered by multiple agencies or law enforcement units. Currently, crash data are being collected by law enforcement electronically at the roadside and on paper tickets that are then keyed into an electronic system. Some law enforcement agencies are able to capture crash data electronically, but nevertheless transmit the information to the DOT in paper form. Due to state budget restraints, limited staff and equipment resources, eligibility require- ments for funding sources, and fears surrounding the confidentiality of personal identifiers, a single elec- tronic data collection system may be difficult to real- ize in the near future. The discrepancies in how data are collected and stored, however, can be managed. Some state DOTs enter into memorandums of understanding (MOUs) with the agency that stores the data regarding how the data will be used and disseminated. Other states have laws that permit DOT access to crash data re- gardless of where the data are stored. Such coopera- tive agreements also are possible between DOTs and MPOs and local planning agencies. For example, one state DOT is developing a training course to educate MPOs on how to use crash data and how to protect the state against litigation. Other states enter into interagency agreements with MPOs allowing them to use the crash data strictly for planning purposes. Importance of Communication In many states where data are not shared among some or all of the planning partners, two major 5

challenges were noted: (1) a lack of communica- tion and (2) a fear of tort liability (as previously dis- cussed). In many instances, a lack of communication prevents MPO access to data. In one state, the DOT representative interviewed indicated that the DOT shares data freely with the MPOs when requested. In a meeting with those MPOs, however, many rep- resentatives said that they did not have access to data nor did they know what type of data the DOT was capable of providing (i.e., raw data, specific intersec- tion or hot spot analysis, summary reports, etc.). Dur- ing a meeting of DOT and MPO officials in another state, the lead engineer for the MPO learned for the first time what types of reports the DOT was able to produce for his jurisdiction. The DOT planning di- rector was shocked by this apparent lack in commu- nication and committed to facilitate a session between the state MPOs and the DOT data analysts. Need for Collaboration Metropolitan planning organizations and local planning authorities can be proactive in gaining ac- cess to crash data, as well as valuable information about safety hazards in their jurisdictions, by asking the following questions: • What department collects and manages the state’s crash data? • Is that department willing and able to share the data with the MPO? If not, why? If so, in what form are the data available? (MPOs may not have the resources to analyze raw data, but many state DOTs provide data in summa- rized form as an additional measure of man- aging risk.) • Are there restrictions on the use or sharing of these data? If so, can the DOT provide guid- ance on how to successfully manage those risks by providing, for example, guidance on the use and protection of the data? Once these questions are answered, the MPO or local planning authority can work with the DOT (or department that manages the data) to identify a process for sharing and protecting the data. There are several strategies states may use to manage risk, although each state’s process may vary due to indi- vidual state laws, jurisprudence, or practice. Identi- fying a process for managing risk may require one or more of the following: • Breaking down barriers to communication between the state DOT (or owner of the data) and the regional, metropolitan, and local plan- ning agencies. More effective communication also may be required between the DOT and the attorney general’s office. • Conducting outreach to all users or potential users (those responsible for planning) of the data to identify their needs and concerns. This may be a logical task for the Traffic Records Coordinating Committee or part of the collab- orative process needed to generate a SHSP. Ask this group of stakeholders to draft a risk management process and share it with the attorney general’s office. • Adopting one of the strategies described in Section 4, such as an interagency agreement, targeted training for users of the data, or de- velopment of an electronic system to track and monitor use of released data. (This approach may be most appropriate for tracking requests by the public.) • Encouraging federal and/or industry organiza- tions, such as FHWA, AASHTO, NHTSA, or the National Association of Attorneys General, to develop a national resource for exchanging information on how states manage the data- sharing risk. Responses to the survey revealed that many states do not know how other states are using or sharing crash data. 4. MANAGING THE RISK OF LIABILITY Many state DOTs recognize the benefits in shar- ing §409-protected data with other agencies, specif- ically other state agencies, MPOs, or consultants under contract to these agencies who are involved in transportation planning activities. DOTs share data with these agencies in a variety of formats, such as crash-specific information, summarized data queries and—in some instances—raw form. Regardless of the format, however, states have developed strate- gies for mitigating the risks associated with sharing these data. Examples of these strategies are provided in Appendix C, with an example of a state Freedom of information Act provided as Appendix D. In all approaches noted in this report, personal identifiers are removed from the data prior to sharing. Impor- tant considerations in managing the risk of liability include: • Immunity, • State liability caps, • Title 23 U.S.C. §409, 6

• Freedom of Information Act, • Tracking the release of information, • Interagency agreements, • Training, • Data format, and • Open data-sharing practice. Immunity Sovereign immunity provides states with effective protection from liability by rendering the government immune from civil suits or criminal prosecution. Many states, however, have significantly lim- ited or even waived their sovereign immunity pro- tection, or have provided a statutory “highway ex- ception” to sovereign immunity.3 Such an exception may provide, for example, that the state govern- mental agency with jurisdiction over any highway must maintain it in reasonable repair so that it is rea- sonably safe. A person injured through the failure to so maintain the highway can sue the governmen- tal agency for damages. There are other various state statutory approaches to waiving or limiting sovereign immunity for highways. These waivers and exceptions can create potential liability for trans- portation agencies on claims against the state for improper facility design and failure to upgrade or improve facilities that appeared to be unsafe based upon crash data. Some state statutes, however, preclude liability for injury or damages caused by the design or plan of a highway, if that design or plan conformed to generally recognized and prevailing standards exist- ing at the time it was prepared. Other states provide absolute immunity from liability for damages result- ing from the design of highways, but there are a num- ber of variations on this statutory “design immunity.” In some states without full sovereign immunity, the liability problem has been solved by statutes which eliminate liability for claims based on failure to per- form governmental functions, such as funding or prioritizing roadway capital improvements. Other states have reached essentially this same result through court decisions on waiver of immunity or separation of powers. A key question for state, regional, and local agen- cies desiring to share crash data is whether their state enjoys full sovereign immunity or any other statu- tory immunity protection, or whether their state has a statutory “highway exception” that may create broad liability for the state highway authority. Immunity status is an extremely important consideration in the willingness to share crash data freely or place con- straints upon its dissemination. State Liability Caps Some states manage the risk of sharing crash data collected pursuant to U.S.C. §152 (and now §148) through the use of state liability caps. These caps put a limit on the amount a plaintiff can seek from a state government agency in litigation. Some states place statutory caps on damages applicable to personal in- jury or wrongful death suits against state and politi- cal subdivisions. States typically set a specific cap amount per suit; however, some state laws may allow higher claims if the local agency being sued has in- dividual caps exceeding the state limit. Two states interviewed noted $300,000 and $500,000 per suit caps. Liability caps may deter some plaintiffs from filing claims against the state; caps, however, pri- marily provide the state with a defined monetary level of protection per awarded claim. The existence of liability caps often influences a state’s willingness to share crash data, with relatively low caps limiting the risk of liability associated with providing data. Conversely, high caps or the absence of caps may serve as a deterrent to the sharing of crash data. Title 23 U.S.C. §409 As described earlier, §409 was enacted to pro- tect from disclosure in litigation data compiled and collected by state DOTs pursuant to Title 23 U.S.C. §152. On January 14, 2003, the U.S. Supreme Court ruled, in Pierce County, Washington v. Guillen, that §409 is a constitutional exercise of Congressional power. (See Appendix A for a more complete analy- sis of the Guillen decision.) Guillen also set out the scope of §409’s protection, clarifying that it only protects data collected pursuant to §152. Section 409, as upheld by Guillen seemingly provides significant protection to states in the proper sharing of data. Many states, however, continue to question whether §409 provides sufficient protec- tion, particularly concerning supposedly protected data that eventually finds its way into the hands of the public or the media through the use of freedom of information or public records act requests, and from there at least indirectly into court. Moreover, some states have interpreted §409 as a “privilege” under which they can then choose to 7

“waive” §409 protection in court. This is done because an FHWA opinion states that §409 does not permit states to use data protected under that statute as a defense. Those states reason that, although waiv- ing §409’s protection permits plaintiffs’ attorneys to use the data against the state, the data are more valuable as a defense for the state. Other states have argued that §409 only limits use of protected data by plaintiffs; the state, therefore, can use the protected data in its defense. It should be noted that FHWA’s position, supported by a strict interpretation of the statute, is that §409 is in effect an “evidentiary bar,” a prohibition that unlike a privilege cannot be waived. Moreover, §409 is a condition of a state’s receiving federal funding and, for that reason as well, cannot be waived. Freedom of Information Acts State freedom of information acts also may limit the protection states desire from §409. Strictly read, §409 is an evidentiary bar, or exclusion, in litiga- tion for damages arising out of an occurrence at a location mentioned in the protected data. Recent decisions from two states’ highest courts support that limited reading of §409 (Telegram Publishing Co., Inc. v. Kansas Dept. of Transportation, Kansas Supreme Court, Case No. 86,767, May 30, 2003; Matter of Newsday, Inc. v. State Department of Transportation, New York Court of Appeals, 3 No. 89, June 9, 2005). Both cases involved a media request, under the state freedom of information act, for §152 data. The state DOT refused to release the data, citing §409. Essentially, both courts ruled that §409 protected such data only in litigation involving the State; that is, its protection did not extend to use by non litigants, such as the media. FHWA has con- curred that §409 has no impact on a state’s duties under its public records laws. A number of states described a tension between their Freedom of Information Act and the protec- tion provided by §409, with some states expressing the view that the existence of state freedom of infor- mation and public records acts significantly limits or even eliminates the effectiveness of the protection provided under §409. This is especially the case given the recent New York Court of Appeals and Kansas Supreme Court rulings. A few states, though, have an exemption from public disclosure for documents not available in discovery. Tracking the Release of Information Many states that share data, manage the risk by carefully tracking the released information. Tracking the release of information is one way of reducing the risk of liability associated with crash information made available in response to a public records re- quest. As one state reported, “The primary benefit of tracking is to allow state lawyers to raise ‘fruit of the poisoned tree’ arguments if a plaintiff and their experts build their case on §409 protected data.” A variety of information tracking systems have been developed. Some of these techniques may be implemented through cooperative agreements among state agencies and with the approval of agency coun- sel or the state’s attorney general. Individual ap- proaches are described below: Public Records Requests—One state has imple- mented a system for sharing location-specific crash data through a Public Records Request. Through this process, interested parties may obtain detailed information about a particular incident through the filing of a Public Records Request. The state’s open records official is able to track these requests and monitor how the state’s data are used; namely to ensure that this information is not submitted as evidence in a legal proceeding. Another state has made its data request form available via the Internet and in printable form for submittal via facsimile or mail. This data request form, in paper form or electronically, includes a field for indicating an agreement not to use the data for litigation purposes against a government agency. If this field is not checked the request is denied, and the requester is sent a letter explaining the state’s rights provided under Federal law 23 U.S.C. §409. Samples of these, and other, tools are included in Appendix C. Watermarks and Stamps—Some states use stamps or watermarks to designate information as protected under 23 U.S.C. §409. These disclaimers serve as official notification of the state’s rights and remind the requester that such data cannot be used against the state in litigation. These stamps may be used on transmittal of raw or analyzed data to either another state agency or in response to a public request. They also may be used in combination with open records request forms or interagency agreements. Digital watermarking is being considered by some states. Digital watermarking technology, used for docu- 8

ments, video, or audio, applies a group of bits to the signal or the author of a signal, which allows tracking of the signal. Through digital watermarks, authors are able to add visible or invisible copyright notices or disclaimers to electronic documents. This tech- nology may be useful as states develop electronic methods for responding to data requests. Software—One state interviewed provides crash data in read-only form via CD-ROM to state con- tracted consultants, along with a disclaimer that the information provided is property of the state. Upon review of the data, the consultant is required to re- turn the CD-ROM to the state. Interagency Agreements One strategy for managing risk involves the use of an interagency agreement or MOU. These agree- ments usually include a description of the protection provided under §409. In many instances, these agree- ments require the requesting party to affirm that their department will only use the data for the approved purposes and will not knowingly allow misuse of the information. Depending upon the agency or organi- zation making the request, the agreement or MOU may require that the requester agrees not to use the crash data for any current, pending, or future litiga- tion against the DOT or another government agency. An interagency agreement is a tool state DOTs may adopt to prevent the misuse of data in litigation. Use of an interagency agreement or MOU be- tween the state DOT or official agency responsible for maintaining the crash record database or re- sponding to data requests and other state, local, or tribal government agencies should be developed and instituted with the advice and consent of the state’s chief legal counsel and other affected state agencies. This strategy allows the state to monitor the requests submitted and creates records of those who have consented not to use such data in litigation against the state. Training Metropolitan planning organizations and local agencies also have responsibilities for transportation safety planning and therefore need access to crash data in their jurisdictions. These agencies, however, also have a significant responsibility to protect from misuse the data that are shared with them. An ap- proach to mitigating the risks of sharing safety data is by training those who need access to the data. State DOTs may consider (a) training the users in MPOs and local agencies on how to properly use the data and how to protect it and (b) educating the staff in these agencies about how the data could poten- tially be used against the state in legal proceedings. In addition, this type of training and communication will help build the working relationships needed to develop a strategic highway safety plan. Data Format Another risk management strategy is for state DOTs to work with a requesting agency to under- stand its technical capabilities. Sending a batch of raw data to an agency or organization that does not have the technical capability to properly analyze and interpret the data may lead to potential misinterpre- tations and misuse of the data—creating potential legal challenges for the state. Small, local agencies often do not have the staff or technical resources to properly interpret crash data. By understanding their “audiences,” a DOT can better serve its planning partners by providing data in formats most suitable to the requester’s needs. One DOT noted that it shares data in summary, raw, and incident-specific forms based on the resources of the requesting agency. Another agency stated that it has a variety of data formats, but always provides detailed analysis and interpretation for any requests made by the media. Open Data-Sharing Practice While some states feel the liability risk of openly sharing safety data outweighs the benefits, other state DOTs and MPOs feel differently and increas- ingly are implementing open, transparent transporta- tion planning and project development processes. Such an open approach, which can be implemented in conjunction with other legal protections, is consis- tent with assessing community impacts and obtain- ing a high level of stakeholder involvement. Towards those objectives, these states approach the sharing of safety data as an extension of their planning ef- forts, particularly in relation to other state, county, or local planning officials. Other states make their data available by posting summaries of crash statis- tics on their websites or through individual open or public records requests. 9

There are a handful of states that proactively share data with the public. For example, one state annually publishes its top 10 high-accident locations. This DOT is able to do this because it asserts the protection provided by §409, denies data requests by litigants within the context of litigation, and has relatively low liability caps. Another state DOT noted that it provides access to crash data because, “access to this information will help everyone do a better job.” One state interviewee noted, when government processes are clearly documented, they often receive judicial support at trial. As a result, the DOT of this state has documented its process for transportation planning, safety planning, and resource allocation. These procedures show how corridors, road projects, and intersections with higher than average crash experience are identified, ranked in priority order beginning with the most dangerous locations, and scheduled for improvement as soon as the neces- sary resources become available. Because the DOT is open with its planning processes and can rely on a certain level of judicial support, it is comfortable publicizing the state’s high-level crash locations and notifying the public of potential roadway safety haz- ards. They take the position that sharing informa- tion with the public may encourage safer behavior by motorists, especially in those potentially danger- ous locations. Another state planning professional indicated that the risk of legal challenge does not outweigh the state’s responsibility to protect the trav- eling public and sharing of this information is con- sistent with an open public planning process. 5. CONCLUSIONS This project identifies how states are success- fully managing the risk of sharing safety data with MPOs and local planning agencies. SAFETEA-LU further raises the visibility of safety as a key plan- ning factor and the importance of data sharing. It re- quires that state, local, and tribal agencies work in concert with MPOs to develop data-driven strategic highway safety plans. Access to crash data will be needed by multiple users for development of statewide goals, objectives, performance measures, strategies, and countermeasures evaluation. As a result, the challenges associated with where data are stored, how they are managed, and by whom will become even greater for the states currently working to man- age risk. Additional measures of liability protec- tion, however, may enable a more collaborative and open approach for statewide and urban area safety planning. Clarification of §409 Protection—Although the Guillen decision answers many questions concern- ing the scope of §409 and its impact on mitigating state risk from tort liability, the decision leaves sev- eral questions unresolved: • A question remains whether §409 will protect data collected by agencies for purposes not re- lated to §148 if those data are first entered by such agencies into an overall system that inte- grates all data entered, regardless of source. By entering data collected pursuant to §148 into, for example, a single, networked, computer- based storage system for crash report infor- mation that also includes data not collected pursuant to §148, can a state use §409’s pro- tection to cover the non-§148 data? • Clarification also is needed concerning whether or not §409-protected data can be used by a state as defense, as well as whether or not §409 can be waived by a state at trial. • Moreover, the interaction of §409 with free- dom of information laws needs to be addressed, particularly after the Kansas Supreme Court and New York Court of Appeals decisions. National Repository for State Highway Tort Claims—The development of a national repository for state highway tort claims would serve as a tool by which states could share information on their defense strategies and thereby further increase their liability protection. While further investigation of the issues identi- fied above is warranted, state DOTs and MPOs can use the practices described in this report to immedi- ately review and improve the manner in which they are managing the risk associated with the sharing of safety data. An important initial step is the develop- ment of an understanding of the Supreme Court’s Guillen decision and the protection provided by §409 and by §148 of the new SAFETEA-LU legislation. This then can be followed by the establishment of effective communication and collaboration with key partner organizations. Building on current laws and court decisions related to immunity, award limits, and availability of public documents and also current policy approaches regarding the manner in which 10

transportation planning and project development is conducted, this review of practices indicates that a number of methods can be utilized, most likely in combination, in conjunction with the protection pro- vided by §409 to adequately manage this risk. This review, however, also indicates that states and MPOs, because of local differences, should not all follow the same approach. The information provided in this report can serve as a guideline for both review- ing current approaches and implementing improved practices. REFERENCES To comply with NCHRP and state requests, no attrib- utable state references are included. (Specific sources were provided to the NCHRP Project 8-54 Panel.) References used for development of this project, in addition to general Internet research, included the following: • Depositions. • United States Code. • State-specific statutes and bills. • Judicial opinions. • Ostensen, G. A. for Federal Highway Admin- istration. Memorandum to Resource Center Managers, Division Administrators. Subject: INFORMATION: 23 U.S.C. 409. March 28, 2003. • Tardif, M. Washington State Attorney General’s Office. Analysis of Pierce County v. Guillen and Its Effect on Sharing of Transportation Data. • Tardif, M. and R. McKenna, Washington State’s 45-Year Experiment in Government Liability, Seattle University Law Review, Volume 29, Num- ber 1, 2005, Seattle, Washington. • Telephone interviews with state DOT planning offices, traffic records divisions, and legal repre- sentatives to state DOTs. • Survey response data (see Appendix B). • West Virginia DOT. 409/FOIA Response Sum- mary. April 2005. ENDNOTES 1. Title 23, U.S.C. §152, the Hazard Elimination Pro- gram, funds activities to resolve safety problems at hazardous locations and sections, and roadway elements which may constitute a danger to motorists, pedestrians, and bicyclists. As further described in Section 2, Congress passed new surface transporta- tion reauthorization legislation after NCHRP project 8-54 began that effectively eliminated §152 and established a new core Highway Safety Improve- ment Program under Title 23, U.S.C., §148. In addi- tion, Title 23, U.S.C., §409 was amended by replac- ing the reference to §152 with §148. 2. To comply with NCHRP and state requests, all attrib- utable state references have been removed from these documents. 3. An excellent discussion of the issues associated with either fully or partially waiving sovereign im- munity is contained in the 2005 article, Washington State’s 45-Year Experiment in Government Liability, by Michael Tardif, Senior Assistant Attorney General in the Torts Division, and Rob McKenna, Washington State Attorney General, published in the Seattle Uni- versity Law Review, Volume 29, Number 1. ACKNOWLEDGMENTS The research reported herein was performed under NCHRP Project 8-54 by Cambridge Systematics, Inc. with support from the American Automobile Associ- ation (AAA). Participants from Cambridge System- atics included John Suhrbier, Principal; Susan Herbel, Ph.D., Senior Associate; and Casey Woodley, Asso- ciate. Susan Pikrallidas, Vice President of Public Af- fairs in the AAA’s National Office, provided legal expertise and assisted in both the interviews and preparation of the final reporting. Members of the project panel were Charles E. Howard, Jr., Transportation Planning Director, Puget Sound Regional Council; Lawrence A. Durant, Gen- eral Counsel, Louisiana DOTD; Gary L. Gittings, Ph.D., Research Associate, Pennsylvania State Uni- versity; Robert J. Shea, Assistant Chief Counsel, Pennsylvania DOT; Jay L. Smith, Assistant Chief Counsel–Risk Management, Missouri DOT; Michael E. Tardif, Senior Assistant Attorney General, Wash- ington State; Raymond Cuprill, J.D., FHWA Liaison; Kenneth Epstein, P.E., FHWA Liaison; and Richard Pain, Ph.D., TRB Liaison. 11

APPENDIX A Summary of Pierce County, Washington v. Guillen The Supreme Court’s decision in Pierce County v. Guillen (537 U.S. 129; 123 S.Ct. 720) is the cul- mination of nearly 40 years of tension between two significant goals: facilitating Americans’ right to a safe transportation system while preserving their fundamental right to redress for negligence. In rul- ing that a Congressional statute, which limits data that can be presented to state and federal courts as evidence in a tort action against states, is a constitu- tional exercise of Congress’ authority, the Supreme Court ended years of uncertainty as to the constitu- tionality and scope of that statute. The statute in question is 23 U.S.C. §409 which states: Notwithstanding any other provision of law, re- ports, surveys, schedules, lists, or data compiled or collected for the purpose of identifying, eval- uating, or planning the safety enhancement of potential accident sites, hazardous roadway con- ditions, or railway-highway crossings, pursuant to sections 130, 144, and 152 of this title or for the purpose of developing any highway safety construction improvement project which may be implemented utilizing Federal-aid highway funds shall not be subject to discovery or admitted into evidence in a Federal or State court proceeding or considered for other purposes in any action for damages arising from any occurrence at a location mentioned or addressed in such reports, surveys, schedules, lists, or data. Legislative History The story of §409 begins in the 1960s when Congress enacted the Highway Safety Act of 1966, which had two principal purposes: (1) to establish national highway safety standards, require states to design programs to implement the standards, and to establish a federal grant program to help states sup- port the programs; and (2) to establish a research and development program. Based on a study conducted by the then Bureau of Public Roads, which resulted in a report com- monly referred to as the Jorgensen Report, the U.S. Department of Transportation in 1967 issued 16 Uni- form Standards for State Highway Safety Programs intended to facilitate state creation of programs to identify and address high-crash locations and to cre- ate a traffic records system, including data about crash locations, crash types, injury descriptions, environ- mental conditions, and other cause and contributing factor information. In 1973, Congress further facilitated the gather- ing of crash data to effect safety improvements by enacting 23 U.S.C. §152. Through this provision, Congress focused federally funded safety improve- ments not only on new interstates, but also—for the first time—on roads which were not part of the federal-aid highway system. The act specifically re- quired that to qualify for federal funding, states must collect and maintain records of highway crash data for use in planning highway safety improvements. Enacted with the best of intentions, §152’s data collection requirements posed a tort liability prob- lem for states. The same detailed and location-coded data on highways, crash causation, and injuries which identified highway locations that could qualify for federal funding for safety improvements also pro- vided rich, detailed, and orderly reports for use by plaintiffs’ attorneys in tort suits for negligence against states for failure to improve “known” dangerous high- way locations. Concerned that, because of liability risks, states would discontinue data collection or be less than thorough in their data collection efforts, Congress in 1987 enacted §409. As originally worded, the statute provided an evidentiary bar for data compiled by states pursuant to §152 “in Federal or State court.” Although a sincere attempt to address the tort liabil- ity problem for states created by §152’s data collec- tion requirements, the deficiency in §409’s protection was soon discovered. First, states found that protect- ing the data from use at trial was not sufficient; the data were being used against them in discovery pro- ceedings. Moreover, since data “compiled” by states for §152 purposes were barred, plaintiffs’ counsel in- stead were demanding the raw data collected pur- suant to §152 for use in litigation. Congress addressed these concerns separately with two amendments: in 1991, Congress made §409 expressly applicable to pretrial discovery; in 1995, Congress added the phrase “or collected” to clarify that the raw data collected by states pursuant to §152 also were subject to the evidentiary bar. It is the 1995 amendment—protecting from dis- closure in litigation data “collected” by states pur- suant to §152—that eventually spawned the lawsuit that the Supreme Court agreed to hear. 12

Background to the Guillen Case Pierce County, Washington v. Guillen began as the result of an automobile crash at an intersection in Pierce County. Prior to the crash, based upon data collected and compiled about that intersection, Pierce County had requested §152 funding to make safety improvements. The request was denied. Pierce County renewed its funding application for the intersection and this request was eventually approved. However, in the meantime, Guillen’s wife was killed in a crash at the intersection. Although the ensuing litigation involved sepa- rate claims, the claim that Supreme Court agreed to hear was based on Pierce County’s refusal to pro- vide, for litigation purposes, documents, requested by Guillen under the State’s Public Disclosure Act, containing information collected and compiled about prior crashes at that intersection. In the deci- sion that led to the U.S. Supreme Court review, the Washington Supreme Court ruled on the scope of Section 409 and on its constitutionality. As sum- marized by the U.S. Supreme Court in Guillen, the Washington Supreme Court reasoned that “Section 409, as amended in 1995, purported to protect from disclosure any documents prepared for state and local purposes, so long as those documents also were collected for Section 152 purposes. In the Court’s view the statute did not turn on the identity of the cus- todian of the document at issue.” (Pierce County v. Guillen, 537 U.S. 129,139; 123 S.Ct. 720, 727.) Stated the Washington Supreme Court: We simply cannot accept the . . . distinction . . . between collections of traffic and accident- related materials and raw data ‘as held by’ Pierce County’s Public Works Department, a local gov- ernment agency involved in ‘section 152 activ- ity,’ and collections of traffic and accident- related materials and raw data ‘as held by’ Pierce County’s sheriff’s Office, which . . . was in no way involved in ‘section 152’ activity. We find such a distinction unsound in principle and un- workable in practice. (Guillen v. Pierce County, 144 Wash.2d 696, 726; 31 P.3d 628, 645.) Having so construed the scope of §409, the Washington Supreme Court then ruled that, in enacting the 1995 amendment protecting from dis- closure raw data “collected” pursuant to §152, Con- gress exceeded its constitutional authority. How- ever, the Court further noted: If this state court has misconstrued the United States Constitution’s limitations upon the Federal government’s power to intrude upon the exercise of state sovereignty in so fundamental an area of law as the determination by state and local courts of the discoverability and admissibility of state and local materials and data relating to traffic and accidents on state and local roads, we are confi- dent that the United States Supreme Court will so instruct, as is its constitutional role under our Federalist system of government. (Guillen v. Pierce County, supra, at 745 and 655.) The U.S. Supreme Court did just that. By grant- ing Pierce County’s writ of certiorari, the Supreme Court took up the case of Pierce County, Washington v. Guillen. Supreme Court Decision Justice Thomas delivered the opinion of the Court on January 14, 2003. In that decision, he first determined the court’s jurisdiction to hear the case. Next, Justice Thomas addressed §409’s proper scope. And, finally, having issued the Court’s ruling on the statute’s scope, he issued the Court’s decision that §409 is indeed a valid exercise of Congress’ authority under the Constitution. Although §409’s constitutionality was a central part of the Supreme Court’s decision in Guillen, it is the Court’s ruling on the scope of §409 which is most critical to the statute’s ability to mitigate state risk from tort liability. In reaching its decision, the Court in particular focused on three views of §409’s scope: (1) that of Pierce County, in which “a docu- ment initially prepared and then held by an agency (here the county sheriff) for purposes unrelated to §152 becomes protected under §409 when a copy of that document is collected by another agency (here the Public Works Department) for purposes of §152” (Guillen, supra, at 143 and 729); (2) that of Guillen, in which §409 protects only materials actually cre- ated by the agency responsible for seeking federal funding for §152 purposes” (Guillen, supra, at 144 and 730); and (3) that of the United States, in which “§409 protects all reports, surveys, schedules, lists, or data actually compiled or collected for §152 pur- poses, but does not protect information that was orig- inally compiled or collected for purposes unrelated to §152 and that currently is being held by the agencies that compiled or collected it, even if the information was at some point “collected” by another agency for §152 purposes. (Guillen, supra, at 144 and 730.) 13

Justice Thomas began by recognizing the ten- sion between construing evidentiary privileges nar- rowly, so as to not impede the search for truth, and the desire to give effect to acts of Congress. The Court then disposed of the first two views of §409’s scope: Guillen’s interpretation of the statute was the narrowest, but it would render the 1995 amendment protecting data “collected” pursuant to §152 “an exercise in futility.” (Guillen, supra, at 145 and 730); Pierce County’s interpretation was determined by the Court to be too broad, “thus conflicting with our rule that, when possible, privileges should be con- strued narrowly.” (Guillen, supra, at 145 and 730.) The Court then adopted the view proposed by the United States as being neither too narrow to give effect to Congress’ intent nor being so broad as to vi- olate the precedent for narrow construction of evi- dentiary privileges. Giving effect to Congressional intent that “§152 was not intended to be an effort- free tool in litigation against state and local govern- ments” (Guillen, supra, at 146 and 731), the inter- pretation protects data collected or compiled by states pursuant to §152, thus addressing “confusion among the lower courts about the proper scope of §409 . . . and [overcoming] judicial reluctance to protect under §409 raw data collected for §152 pur- poses.” (Guillen, supra, at 146 and 731.) On the other hand, the interpretation narrowly construes the priv- ilege by making plaintiffs no worse off than if §152 funding never existed. Said the Court, “. . . there is no reason to interpret §409 as prohibiting the disclo- sure of information compiled or collected for pur- poses unrelated to §152, held by government agen- cies not involved in administering §152, if, before §152 was adopted, plaintiffs would have been free to obtain such information from those very agencies.” (Guillen, supra, at 146 and 731.) Having settled §409’s scope, the Court then found the statute to be a constitutional exercise of Congress’ authority. Although Guillen had chal- lenged the statute’s constitutionality under the Con- stitution’s Commerce, Spending, and Necessary and Proper Clauses, because the Court concluded Con- gress had authority under the Commerce Clause, it did not need to reach the challenges under the other Clauses. APPENDIXES B THROUGH E Appendixes B through E of the research agency’s final report are published as NCHRP Web-Only Doc- ument 89 (see Online Documents at www.TRB.org). These appendixes are titled as follows: Appendix B: Survey of States Appendix C: Examples of State Risk Manage- ment Practices Appendix D: Example of a State Freedom of Information Act Appendix E: Example from International Safety Data Scan 14

Transportation Research Board 500 Fifth Street, NW Washington, DC 20001 These digests are issued in order to increase awareness of research results emanating from projects in the Cooperative Research Programs (CRP). Persons wanting to pursue the project subject matter in greater depth should contact the CRP Staff, Transportation Research Board of the National Academies, 500 Fifth Street, NW, Washington, DC 20001. COPYRIGHT PERMISSION Authors herein are responsible for the authenticity of their materials and for obtaining written permissions from publishers or persons who own the copyright to any previously published or copyrighted material used herein. Cooperative Research Programs (CRP) grants permission to reproduce material in this publication for classroom and not-for-profit purposes. Permission is given with the understanding that none of the material will be used to imply TRB, AASHTO, FAA, FHWA, FMCSA, FTA, or Transit Development Corporation endorsement of a particular product, method, or practice. It is expected that those reproducing the material in this document for educational and not-for-profit uses will give appropriate acknowledgment of the source of any reprinted or reproduced material. For other uses of the material, request permission from CRP.