This document, without its Appendices, may also be downloaded as a Microsoft® Word file (about 1/2 megabyte in size).

EXECUTIVE SUMMARY

As the primary source of basic data on travel conditions, public agencies through their data sharing practices can have a powerful effect on deployment of 511 telephone numbers and other types of traveler information services. This report documents the current state of the practice, describing how the public and private sectors deal with data ownership and sharing, and examines policies aimed at facilitating data sharing and ultimately improving the quantity and quality of information that reaches travelers.

The report is based on information collected from two sources. Surveys were conducted with thirty-four public agencies and seven private firms. The surveys consisted of interviews with representatives of public and private sector entities that are active participants in data sharing. The 30-minute interviews included a variety of questions about data sharing practices, such as the types of information shared, recipients of the data, and types of conditions placed on users of the data. The surveys were complemented with a review of the literature about data sharing practices related to traveler information and other types of data.

Major findings of the research are highlighted below. It should be noted that the use of the term "data" in the study encompasses digital, video, and verbal forms of information.

• Agencies have two major objectives in sharing their data with private sector and other public sector recipients: improving transportation operations through better interagency coordination and optimizing the use of the transportation system by providing information to travelers. Enhancing interagency coordination was the top-ranked motive for data sharing.
• Even though their motives are different, public and private sectors are active participants in use of traveler information as a transportation management tool. Almost all agencies directly provide information to the public typically with VMS, HAR, kiosks, and interactive voice response telephones. Although agency data are a fundamental source, private providers generally need to enhance public data before they are marketable. The most common types of information provided are traffic and road conditions, incident information, and planned construction information. Transit data are generally less useful to private providers, and only a third of them report transit delay information.
• Agencies who have data to share protect their interests by placing restrictions on access to data, but firms generally do not find these conditions to be onerous. Two or more conditions on access are common, the most frequent being acknowledgement of the agency as the source of the data when distributed to the public.
• Formal policies on data sharing were reported by half the surveyed agencies and several more have plans to issue one. The principal advantage of a formal policy is that it provides a process for handling requests for agency data.
• In addressing the costs associated with the data sharing process, agencies frequently employ two or more cost recovery mechanisms in data sharing relationships. Most frequently agencies require the receiving party to cover its own cost, such as hardware, software and communications cost to connect to agency data sources. The second most popular mechanism involves a private firm sharing its "value-added" information with the agency.
• The two most controversial topics in the private sector's relationships with agencies regarding agency data are revenue sharing and exclusivity.
• The idea of revenue sharing is optimistically viewed by many agencies, although in practice it has not had much success. The private sector tends to oppose revenue sharing either because of practical difficulties in administering it or because it violates the principle that public data should be available to all taxpayers for free.
• Exclusivity is the model used by only about 15% of the surveyed agencies, which see its value in assigning to some other entity the burden of dissemination of agency data. Private firms are generally opposed to exclusive arrangements because they constitute monopolistic, anti-competitive franchises and because they violate the principle of right of access to data collected at taxpayer expense.

The data sharing policy of the National Weather Service (NWS) was examined. The NWS minimizes the control it exercises on its data used by others and does not seek to profit in the dissemination. The economic benefits of the booming private sector weather information business are seen as validation of the NWS policy. The data sharing practices of the NWS could serve as a useful model for transportation agencies which generate data in the course of performing their planning and operations functions and, at the same time, share data with private entities to create economic benefits.

The FCC ruling on 511 signaled a new phase of traveler information services and drew new attention to the role of transportation agencies. The attention is new because the FCC assigned control of use of the 511 number designation to state and local agencies. However, transportation agencies have long held a key role as the principal source of information about travel conditions, even when other parties deliver that information to travelers.

The act of data sharing by transportation agencies is the subject of this report. Ownership and rights to use could have a powerful effect on deployment of 511 and other traveler information services. Thus, the report documents the current state of the practice, describing how the public and private sectors deal with ownership and sharing. The report also examines policies aimed at facilitating data sharing and ultimately improving the quantity and quality of information that reaches travelers.

The scope of the study commissioned by U.S. DOT included both the agencies who generate and disseminate data as well as the recipients, especially private sector firms, that provide traveler information services to travelers. The objective was to gather information on several aspects of data sharing practices, including:

• types of information being shared and the formats of their transmission
• conditions placed on using the information
• technical and business arrangements established between the sharing parties
• policy support for data sharing.

The remaining sections of this report describe the methodology used in the study (Section 2.0), the results of the literature review and survey (Sections 3.0 through 5.0), policy issues (Section 6.0), and conclusions based on the findings of the study (Section 7).

2.0 APPROACH

The study adopted a two-pronged approach to the subject of data sharing: a literature review and a survey of active participants in data sharing.

2.1 Literature Review

A review of literature provided an opportunity to learn from earlier studies that had examined the topic of data sharing. The review focused on identifying and documenting findings from traveler information studies, but it also searched for other transportation and non-transportation sources to understand how data ownership was being managed.

Relevant literature was sought from a number of sources. An on-line search was conducted using Websites of the Transportation Research Board (TRB), National Technical Information Service (NTIS), Transportation Research Information System (TRIS) Database, ITS America Web page, U.S. DOT's Electronic Document Library (EDL), and the ICDN. In addition a search of CD-ROM-based proceedings from meetings of ITS America, the ITS World Congress, and the TRB was conducted. Over 30 articles were identified that matched the key words associated with the topic of the study, and of these 11 were deemed relevant.

To locate literature on data sharing outside the domain of traveler information systems, an informal network of contacts was queried. Several potentially fruitful sources were identified and investigated. The most relevant of these were the data-sharing practices of the National Weather Service, which is reported on in Section 3.0.

2.2 Survey Method

The range of public agencies' data sharing practices relative to traveler information services was investigated in a survey of agencies currently sharing data. A list of candidate agencies for the survey was assembled using systems identified as having traveler information telephone numbers (Schuman et al. 2001). That list was supplemented with names of agencies that have traveler information projects or services that were known to the authors of this report. The candidate agencies were prioritized based on criteria such as regional representation and types of data being shared.

To provide a complete view of the nature of data sharing practices, the perspective of private sector firms that receive public data was also sought. A list of candidate firms involved in traveler information services at a local or national level was assembled.

A set of six "scoping interviews" was conducted in March 2001 with public and private sector representatives chosen for their in-depth and relevant experience with data sharing in the traveler information arena. Lasting from 30-90 minutes in length, the scoping interviews provided valuable insight into the dimensions of the data-sharing topic that deserved more exploration. The next step was the development of separate public sector and private sector questionnaires to be administered by telephone.

A total of 34 telephone interviews were conducted with the public sector, exclusive of the scoping interviews, during April and May of 2001.[1] A total of seven interviews with private firms were conducted. Three of these occurred in June of 2001 and the remainder consisted of private sector responses from the earlier scoping interviews, which the study team decided to use to increase the overall sample size for the analysis. Appendix 1 presents a list of agencies and firms interviewed.

A note of caution is in order regarding the sample. The agencies and firms interviewed were selected because they were known to be involved in data sharing and, in some cases, are leaders in that regard. The sample illustrates the range of experiences and opinions among these leaders, but the sample is not truly representative in the statistical sense. Thus, no claims are made about the statistical significance of the results in this report. In addition, DOTs are heavily represented among the public agency respondents, and caution should be used in extrapolating the results to all types of transportation agencies.

3.0 FINDINGS FROM THE LITERATURE

The traveler information literature reviewed for this report yielded only modest amounts of relevant findings, but the literature commercialization of public information outside of transportation proved instructive. A bibliography at the end of this report identifies all the sources examined. The sections below summarize the findings from both transportation and non-transportation literature.

3.1 Traveler Information Literature

While data ownership and data sharing are not frequently addressed topics within the literature of traveler information systems, issues surrounding these topics occasionally can be found. The most pertinent literature is cited here.

In a case study of the TransGuide Metropolitan Model Deployment Initiative in San Antonio (Booz-Allen, May 2000), many institutional issues that arose in the MMDI revolved around legal and policy constraints placed on publicly funded projects and property. With respect to sharing of data, the policy in place at TXDOT at the time was to share as much information as possible, a reflection of the Texas Open Records Act, which made all transit and traffic signal data readily available to the public with few restrictions. To formalize the intention to share data, Texas Transportation Institute drafted a policy on behalf of TXDOT that would allow access over the telephone or Internet of traffic speed data and video images. Thus, as the TXDOT example illustrates, a general state statute can provide the spirit or basis for an agency's policy for sharing of traveler information.

Ownership and control of data was a key aspect of ATIS business models that have been discussed in several citations: Hallenbeck et al. (1998), the companion document published by ITS America (1998), and workshop proceedings on ATIS data collection guidelines (2000). Although the authors acknowledge that many variations on the five general models are possible, nevertheless, the models serve to differentiate modes of business relationships for ATIS. At one extreme is the model emphasizing public sector interests, with the public sector generating, fusing and disseminating data and retaining a high level of control and ownership of the data. The openness of access for both the general public and commercial service providers tends to limit the private sector's opportunities for revenue generation. The franchising model represents the other extreme with respect to data ownership and control. In this model, the private sector relies on the public sector only for data collection and a private firm performs fusion, dissemination, and operation of the traveler information service. Because the private sector acquires the rights to market the data for a given length of time, the public sector loses control over ownership of the data as a trade off for expectations of cost savings and superior dissemination to travelers. Thus, as the franchise model suggests, public agencies may adopt business models in which control over data is bargained for other benefits that the private sector can offer.

The literature search uncovered several cases of formal statements of policy or contractual language dealing with data sharing for traveler information. Moreover, several policy statements or contractual agreements from agencies around the U.S. are included in a report by Clinger and Abedon (2001) that focuses on the issue of privacy surrounding dissemination of public sector CCTV images. Policy statements were also obtained from a number of the public agency representatives interviewed during the survey portion of this report. Appendix 2 contains copies of policy statements and/or contractual language from the following entities:

• Virginia Department of Transportation
• PB Farradyne on behalf of the Metropolitan Transportation Commission, Oakland, California, and other TravInfo® partner agencies
• Las Vegas (Nevada) Area Traffic Computer System
• Texas Department of Transportation
• Ohio Department of Transportation and Kentucky Transportation Cabinet
• Road Commission for Oakland County
• South Carolina Department of Transportation
• Missouri Department of Transportation
• Illinois Department of Transportation
• State of Connecticut

The documents in Appendix 2 deal with the relationship between the public agency generating the data (including video images) and private entities receiving the data. Four of the documents are general statements of policy, and six are contractual agreements many of which embody statements of policy. Three of the documents pertain to video images only, three deal with electronic data only, and 4 deal with both forms of data.

The types of issues pertinent to data sharing addressed in one or more of these documents include:

• Type of public data to which private entity has access
• Exclusivity (or non-) of access to data
• Fees or costs borne by private sector associated with access and dissemination of the public data
• Sharing of privately generated data with the public agency and protection of its value
• Dissemination of data by private entity to other parties (free or resale)
• Control over video images
• Requirements that the private entity use the public data to provide information to travelers
• Acknowledgement of public agency as source of data (e.g. logo)
• Technical requirements for access, e.g. communication system
• Allowable advertising
• Personal identification and graphic images of serious injuries
• In-kind services provided by private sector (e.g. airtime for public service announcements)
• Training on use of the public data
• Intellectual property rights
• Standardization of format of data
• Monitoring of usage of traveler information services
• Liability for data quality and availability
• Sharing of private entity revenue with public agency
• Placement of private equipment on public rights-of-way
• Raising public awareness about traveler information services

Variation among the policy statements and contractual agreements can be observed on the specific issues they address, and the position taken by an agency on an individual issue. For example, some agencies want to maintain knowledge about and control over third parties' use of the data, whereas other agencies appear to encourage wide dissemination and use without close scrutiny. The one general observation that can be made has to do with cost. In every case the public agency requires the private sector to assume all costs for its equipment, software, and communications to access the public data.

3.2 Other Literature

Although time and resources did not permit a thorough examination of literature on data sharing practices and issues in other transportation and non-transportation areas, the practices of the National Weather Service (NWS) proved extremely relevant to the topic of this report. Data sharing in other applications, including geographic databases, data on motor vehicle and drivers licensing, and satellite imagery were investigated. They did not appear promising for the purposes of this study and were not further explored. Thus, the policy of the NWS, which was the most pertinent source, is discussed in detail below.

Data Policy of the National Weather Service

Data sharing at the NWS is discussed by Weiss and Backlund (1997) and Kelly (nd). Underlying NWS' approach is the view that good weather data are valuable because they reduce the monetary and non-monetary costs of weather events for businesses and individuals and provide raw material for profit-making businesses, such as those that supply weather related information to the public. The highly successful Weather Channel on cable television is an example of the benefits of making data freely available to the private sector. The parallels for transportation data are the potential for reducing costs associated with transportation congestion and delays and the potential for private transportation information companies.

The policy of the NWS is based on the Federal policy of open and unrestricted access to taxpayer-funded government information. Federal policy flows from the idea that the way to maximize the benefits of data is to provide them at the cost of dissemination to as wide an audience as possible. Federal laws and regulations from which NWS derives its data sharing practices are:

• The Copyright Act (17 USC. 105)
• Freedom of Information Act (5 USC 552)
• Paperwork Reduction Act (44 USC Ch. 35)
• OMB Circular No. A-130 (61 FR 6428, February 20, 1996)
• Electronic FOIA Amendments Act of 1996
• Government Paperwork Elimination Act (PL 105-277, Div C, Title XVII).

One of the most important is the OMB Circular No. A-130, which states that Federal agencies should:

• Actively disseminate all public information
• Provide data without restrictions or conditions
• Provide data at no more than the cost of dissemination
• Take advantage of private, academic, and other channels of dissemination
• Use the best available technologies, such as the Internet, satellite downcast, etc.

Consequently, the National Weather Service provides data in a variety of different ways to a variety of audiences, from simple to sophisticated. An example of NWS data sharing can be seen in NEXRAD radar data, which is one component of NWS products. NEXRAD provides data for basic NWS forecasts, watches, and warnings, but NWS does not keep NEXRAD data to itself. NWS wants people to have the data to maximize the return on taxpayers' investment in the system and to provide the data to the widest set of users possible. NWS has three different and complementary internet-based technologies and approaches to provide real-time access to NEXRAD radar data and products:

• Real-time data stream to anyone choosing to bear the cost of a connection and direct dissemination costs (power users)
• Anyone can select and download for free any product by Internet file transfer protocol (FTP) (mid-level user)
• Four basic radar images available for viewing on the Web (basic users).

With respect to the costs of obtaining NWS data, a user pays for transmission costs only if NWS has to recover any costs incurred in setting up the dissemination mechanism to feed data to the user. No attempt is made by NWS to profit on its data, because they view their relationship with the private sector as a strategic partnership. That is, the public should be able to take advantage of many channels for data dissemination, including private for-profit, private not-for-profit, academic, etc. The NWS discourages government giving exclusive rights to a private organization for a fee in the hopes that the revenue produced can be used to make data more accessible. They argue that such monopolistic arrangements dampen competition hurting data production and dissemination in the long run.

In the area of controlling how the data are used, NWS relinquishes all control once it leaves NWS with a couple of exceptions. They do not allow data to be modified and then presented as official government material. Moreover, users producing copyrighted materials based on data and/or products from the NWS must provide notice identifying the NWS material and state that such material is not subject to copyright protection.

Liability is handled in a straightforward manner: the NWS does not assume any liability with regard to the use of the data. They caution users to be aware of the date and time of the data and products being used and state that NWS provides the data "as is." The user assumes any risk related to their use, and "in no event is NWS liable to you or anyone else due to your use of these data."[2]

Can Public Agencies Successfully Commercialize Their Data? Comparing NWS with Other Agencies

Because of their own policy and successful experience in encouraging businesses to take government information and customize it to add value for customers, NWS argues that government should not be in the commercialization business. Commercialization is the practice of a government agency charging for public information that was financed by general tax revenue - weather, geographic, etc. This is not to be confused with privatization, in which functions not inherently governmental, such as air traffic control and telecommunications for example, are transferred to the private sector. Another premise of the NWS viewpoint is that the government should not duplicate what the private sector is doing.

The experience of other agencies, both in the U.S. and abroad, attempting to profit on the data they generate, raises doubt about the true potential of such ventures. The following problems have been identified by Weiss and Backlund (1997):

• Shell-game practice. Government commercialization has not worked because often the organizations using the data are other public agencies; therefore, financing is a shell game with the government budget and does not raise revenues from outside government. For example, the British Meteorological Office gets 70 percent of its revenue from the Ministry of Defense.



• High elasticity of demand for information. Attempts at commercialization can lead to prices charged to the private sector that many view as too high. The result is that demand for the data falls off very quickly and revenue to the public agency is usually very weak. Information has high elasticity of demand because people will often do without.



• Leakage of information. Information can be shared with no loss to the sharer (it is non-exclusive) making it difficult for an agency to enforce copyright.



• Barriers to entry and cream skimming of markets. When the price charged for the data reaches the cost to independently replicate, private vendors will do so and they will start with the most lucrative markets (cream-skimming). Examples include the Federal Maritime Commission's unsuccessful attempt to charge for tariff information, and the British private mapmakers publication of city maps leaving to the government mapmakers the unprofitable but necessary task of mapping the remainder of the country.

Summary of Findings from Literature Review

The literature on traveler information has shown that aspects of data sharing are being addressed as both an abstract concept and a practical problem. As abstractions, the ATIS business models characterize public and private sector roles around the ownership and control of traveler information data. Agencies adopt the model that best suits their needs and resources, and, in the process, determine the extent of control they exercise over the use of publicly generated data. At the practical level public agencies are codifying their positions on control of public data used for traveler information in statements of policy and language in contracts with private sector recipients.

Interesting lessons for traveler information with regard to data sharing were found in literature about the practices of the National Weather Service and other public agencies' attempts at commercialization of public data. NWS seeks wide dissemination of its data to public and private users. NWS minimizes the control it exercises on its data used by others and does not seek to profit in the dissemination. In contrast to NWS practice of free access, many attempts at commercialization of public data in the U.S. and other countries have not fared well. Information is inherently difficult to control and cannot easily be priced to reflect the true cost of collection. Similar issues exist in the domain of traveler information, and the ATIS community can benefit from the lessons offered by NWS and other government entities.

4.0 CHARACTERISTICS OF RESPONDING AGENCIES AND FIRMS

This section describes the surveyed agencies and firms. First, some basic demographic characteristics of the responding agencies are presented. Next the public agencies' motivation for sharing data is discussed, followed by the types of traveler information services that the respondents provide. The fourth section presents data on the business models that the surveyed firms employ. In the final section the private firms offer their opinions on what does and does not work in their arrangements with public agencies. Throughout the report a shorthand notation is used to refer to questions in the public and private sector surveys. Public sector questions are noted as PubQx, where "x" refers to the question number in the questionnaire (see Appendix 3). Similarly, the private sector questions are noted as PvtQx (refer to Appendix 4).

4.1 Demographics

Representatives from thirty-four public agencies were interviewed about the types of data they collect and share with other organizations. Another seven interviews were conducted with officials of private sector organizations about their experiences with receiving data from the public sector. Six of the seven private organizations are for-profit and one, the Travel Advisory News Network (TANN), is a private company created and operated by The Partnership, a non-profit public/private partnership. TANN acts as an intermediary between public agencies and private companies in the Los Angeles region. All of the organizations interviewed currently collect and share data used in traveler information systems or provide data in a traveler information system.

Twenty-eight of the public agencies interviewed are concerned with traffic or multi-modal data, three with road condition data, and three deal with transit data. About two-thirds of the agencies provide metropolitan level travel data, with the remaining one-third providing information at the state or regional level. The responding public agencies are geographically diverse with 10 from the West, 9 from the Midwest, 11 from the South, and 4 from the Northeast. They include state departments of transportation, regional agencies, and transportation agencies within local government.

Most of the businesses interviewed are involved in providing traveler information in more than one metropolitan area/state. Moreover, most appear to be working towards building a network of information systems nationwide. Two, an NBC television network affiliate and the Travel Advisory News Network, focus on one metropolitan area each.

4.2 Motivation for public data sharing

Most agencies are motivated to share public travel data to enhance coordination among the region's transportation agencies to improve overall travel conditions (PubQ4). Nineteen of 34 agencies ranked this as the number one reason to share data when asked to prioritize their motivations (table 1). In many cases this stems from an internal policy to use traveler information to improve the utilization of the transportation system, the second highest ranked reason. A regional program requiring public agencies to share data was deemed less important by many of the public agencies. Only a few organizations cited solicitations from the public and private sector as the main reason for data sharing.

Table 1. Public Agency reasons for Data Sharing
(Prioritized from 1 = most important to 5 = least important)

4.3 Types of traveler information services delivered by public or private sector

Figure 1 shows the types of traveler information services that respondents said are provided by the private sector in their area (PubQ15). The types of traveler information services most commonly provided are traffic and road conditions and incident delay. Somewhat less frequently provided is planned construction and special event information. Transit delay information is provided in about a third of the areas covered by the respondents to the public agency survey. Five areas reported the provision of other types of information including weather and air quality information, alternative routes, parking, public safety campaign information, and long distance trip information.

Almost all the public agencies currently offer information directly to the public or are planning to in the near future (PubQ14). Web pages are the most common form of dissemination followed by VMS/CMS, kiosks, HAR, and interactive voice response telephone. Operator assisted telephone information was the least cited, but still eleven public agency respondents mentioned this. Twelve agencies mentioned some other way of disseminating information directly to the public, such as pagers, PDAs, cable TV, e-mail, and fax.

4.4 Private sector business model

Typically, businesses collect public data, supplement it with private data, and then sell it to other businesses or the general public (PvtQ1). Data is collected privately in many ways including direct observation from an airplane, information provided by cell-phone users, and the monitoring of police scanners. The types of businesses supplied with the fused travel information include fleet management companies, wireless and broadcast carriers, and sellers of traffic information services and devices to the public. Some businesses, particularly TV stations, use traffic information as news content with revenue derived from advertising sales. Some businesses have more complex relationships with public agencies. One business in the survey is paid by a public agency to run its traffic management center. Another operates and maintains publicly funded ITS infrastructure in the public right-of-way.

4.5 Private sector opinions about workable and unworkable arrangements with public sector

The views of the private sector on the arrangements between the public and private sector agree, to a large extent (PvtQ3 and PvtQ4). Most feel that public sector agencies should make data available to all without a fee over and above the cost of dissemination. Many respondents specifically mentioned the view that public agencies should avoid exclusive data sharing arrangements.

One respondent thinks that public agencies should avoid revenue-sharing agreements, partly because government often does not know how to accomplish it successfully. Many expressed the view that it is usually possible for businesses to share value-added data with the public sector, especially if accompanied by an agreement that the data will not be shared with other parties. One respondent said that public agencies could provide private sector access to the public right-of-way for data gathering. And another argued that the government should stay away from trying to commercialize travel data itself.

5.0 PRACTICES OF PUBLIC DATA SHARING: PUBLIC AND PRIVATE EXPERIENCE

In this section of the report the current practices and perceptions of public and private participants in data sharing are discussed. First, the types of data shared are discussed. Next the requirements that agencies place on sharing data is presented followed by the formal and informal policies that have developed. Cost and revenue considerations are discussed in the fourth section. Finally, lessons learned by the participants in data sharing activities are presented.

5.1 Types of Data Shared and the Recipients

To understand the domain of data sharing practices, public agency respondents were asked about the types of transportation data that they share and who receives it. For the purposes of the survey, real-time was defined as data on current conditions on the roads or transit systems, such as traffic speeds or transit delays. Static information was defined as data that change infrequently, such as road construction schedules and bus schedules.

What Data Do Public Agencies Share?

Agencies most frequently share highway-related data and real-time highway data are the most common type of information (PubQ3). As illustrated in Figures 2 and 3 electronic/digital form is the most popular (24-25 of the 34 agencies), although verbal and video forms are only slightly less common.

Whether the recipient is a private sector firm or another public agency did not make much of a difference in the type and form of data shared. About the same number of agency respondents said they share the three types of data with other agencies as the number supplying the private sector.

Static highway information in electronic/digital form is provided by somewhat fewer agencies compared to real-time data. For instance, for the private sector twenty agencies provide static highway electronic/digital information and fourteen verbally transmit static information. Not surprisingly, video transmission of such data is rare.

Transit agencies do share data, most often with other public agencies. Of the three exclusively transit agencies all provide static transit information in electronic/digital form to other public organizations, and two provide real-time data in that form to the public sector. Private sector organizations are less involved in receiving transit information, particularly verbal and video data.

Figure 2 represents public agencies' views of what they are sharing with private firms. When asked what they receive, the seven private sector respondents generally confirm this pattern of data sharing. Highway real-time data are either typically or occasionally received in all forms at the locations in which the firms operate. Similarly, static highway data are received by the private firms, but slightly less frequently than real-time data. Transit data in any form, on the other hand, was never received by some of the private firms. These findings clearly reflect the current dominant product in the private traveler information market: real-time traffic conditions.

Who Receives Public-Sector Data?

When asked to identify the types of parties with whom they share data (PubQ2), agencies reported that other public agencies are the most frequent recipients of data (Figure 4). Of the 33 agencies that answered this question, 31 share data with other public agencies. The category "other public agencies" is followed by, in order of frequency mentioned, local TV, traffic reporting organizations, local radio, Internet service providers, other organizations, and local newspapers. About a third of the data providers supply local newspapers with information.

Among public sector organizations the most frequently cited were other local jurisdictions such as counties and cities and more specific departments such as the department of public works. Other organizations frequently mentioned include the state police, 911 systems, the State DOT, and transit agencies. Mentioned less frequently were emergency management departments, an airport, a university, and a state parks agency.

A host of other organizations were also mentioned as data recipients. These included major traffic generators by one respondent, such as malls and large businesses, and another mentioned the New York Yankees. Others include a realtor company, an in-vehicle content provider, and a data integrator.

5.2 Conditions Placed on Data Sharing

Making data widely available to public and private sector recipients does not mean that no strings are attached (PubQ11). Twenty-eight agencies report that they place some type of condition on accessing the data. Six have no conditions. Of those with conditions twelve agencies place three conditions on the data user, with seven each with one or two conditions. Only one agency had four conditions. Figure 5 shows the type of conditions placed on data sharing.

Respondents, when asked to elaborate on the nature of the conditions that they impose on recipients, made the following comments:

• Fourteen place conditions on the hardware and software required to access the data. A common comment is that the user needs specific hardware and software to be able to use the video data.
• Eighteen data providers place conditions on the use of the data, with most of these restricting pan, tilt, and zoom of video to prevent identification of license plates and crash victims.
• Twenty-five data providers require acknowledgement of the data source. Typically, the data user needs to give both a verbal and a written acknowledgement (in the form of a logo and/or the source agency's name), although this often depends on the type of data being used and the type of user (radio, TV, internet, etc.).
• Four agencies have other types of condition. These include agreeing to leave the data unmodified, not using the data for enforcement, and not reselling data.

With the public sector making conditions part of the bargain, the private sector respondents were able to provide their perspective on the effect of the conditions. In general, they do not appear to be onerous. All firms reported the need to meet agencies' technical specifications for receiving the data. Most firms also report having to comply with restrictions on use, such as not identifying crash victims and license plates. One firm reported that there was a company policy in this area, even when an agency did not impose a restriction. Concern was expressed over not having control over the camera, a disadvantage when an agency is not managing cameras with traveler information in mind. Mention was also made about an instance of a transit agency not allowing a firm to change information on delays, even when the data were incorrect. Acknowledgement of the public agency source, especially on video images, is common. However, some firms feel acknowledgement is positive as it provides credibility to the traveler information being disseminated. A couple of firms indicated that they had encountered other conditions on use of public data. In one location a state DOT did not want locations identified that have a lot of accidents for fear of law suits that might ensue. This concern had to do with historical data not real-time data. Another condition reported by one firm was that public data were available only to locally based firms, a rather discriminatory restriction, it would seem.

When it comes to specific methods used to retrieve data, twenty-nine of the thirty-three agencies that responded to the question cited a direct connection of some sort for receiving the data (PubQ12). The next most cited methods are the Internet (20), fax (14) and e-mail (10). Other methods by frequency of occurrence include phone, radio, bulk mailing, pager, a kiosk with public Internet access, and television.

Agencies were asked if they had some type of limit on the number of entities that can simultaneously access system data (PubQ13). Of the sixteen agencies responding affirmatively, most cited limits to the equipment they had in place as the cause. The most severe limitation seems to apply to video, although the dial-up data system at one public agency is limited to five users at a time.

5.3 Agency Policies on Data Sharing

Formal Policy on Data Sharing

As revealed in the literature review (Section 3.0), agencies have established policies regarding data sharing in some cases. As shown in Figure 6, just over half of the agencies surveyed have a formal policy about sharing data (18 of 34) (PubQ5). Only 9 of the 18 knew when their policy was issued, although another four referred to the enactment of a statewide policy. Of the known dates seven were issued relatively recently (since 1990), with 1 each issued in the 1970s and 1980s.

Twelve of the eighteen agencies with a policy said it was established to provide a process for handling requests being made for agency data. Ten of the eighteen said the agency has a policy to help disseminate traveler information. Eight agencies mentioned some other reason including: to facilitate agency coordination for incidents and construction; to ensure fair treatment; to deal with privacy issues; to establish public/private partnerships; and to manage expectations and resources.

In the absence of a formal policy, an agency can still express its views on data sharing through other means (PubQ6). Nine of the sixteen agencies expressed views on data sharing in contracts or agreements. Two others incorporate data sharing views in training or procedure manuals.

Six agencies have plans to issue a formal policy, and have several reasons for doing so. Five are motivated by a desire to help disseminate traveler information, and the same number want to establish a process for handling requests being made for agency data. One agency reported that the Federal requirement for an ITS regional architecture has stimulated their interest in a data sharing policy

Eight agencies stated that they had no plans to develop a policy. Six of these said they did not need a formal policy because they have no problems. Other reasons mentioned are that agreements form the policy, that there is no demand for data, and that existing state law prohibits the development of a policy.

Use of Written Agreements

Seventeen out of thirty-three agency respondents reported that written agreements are always required when they share data with other parties (PubQ7). Another twelve said they are needed sometimes, with the most often cited reasons being when data are shared with a private organization, when video data are involved, and when there is a charge for the data. Four agencies reported that written agreements are never required.

The experience of the private sector recipients tends to mirror the practices being reported by the public sector providers of data (PvtQ10). For video and digital information, respondents said a written agreement was required all the time or sometimes. On the other hand, verbally transmitted data sometimes required a written agreement but for some respondents a formal agreement was never needed for verbal data.

5.4 Costs and Revenue Considerations

Cost recovery mechanisms

Cost can be an important issue for both agencies sharing data and recipient organizations. For one or both parties labor, equipment, and communication costs can be involved in making the relationship work from a technical and institutional standpoint. As the supplier of the data, public agencies may attempt to protect their financial exposure or recoup their costs in various ways as the survey results reveal in Figure 7 (PubQ10).

Most commonly, agencies require users to pay for their own hardware, software, or communications lines to access data (28 out of 34). Direct reimbursement by the user is much less common -- just eight agencies reporting it. Although only seven agencies receive some type of in-kind contribution, fourteen agencies require the receiving party to make their value-added information available to the agency for internal use. Finally, seven said that the user is required to share a portion of the revenue generated from its business. The subject of revenue sharing is examined in more detail below.

The largest number of agencies (12 of 34) has two types of arrangements. Six of the twelve require data users to pay for their own equipment and to make value-added information available to them for internal use. Nine agencies have three types of arrangements, and two have four arrangements. Six agencies have only one type of arrangement and five have none at all.

The private sector's experience with agency cost recovery practices is consistent overall with what the agencies report (PvtQ8):

• The private sector typically (5 out of 7 respondents) pays for its own hardware, software, and communication costs for accessing agency data.
• Five respondents indicated that they typically (2) or occasionally (3) provided their "e;value added" information to the public agencies.
• Four firms said they made in-kind contributions occasionally (3) or typically (1).
• Revenue sharing is done by three of the firms (1 typically, 2 occasionally).
• Direct cost reimbursement to the agency was the least common practice and was reported as occasional by only 2 firms.

Opposing views on revenue sharing

On the subject of revenue sharing, eleven agencies reported that they have sought to recover revenue from the use of their data beyond the costs of making it accessible (PubQ8). These eleven agencies appear to be inconsistent with the seven agencies previously mentioned that said they require sharing of revenue generated from the use of the data. The explanation might have to do with attempts that were tried but were not viewed as successful and, therefore, the requirement was dropped. The hoped-for success of revenue sharing was not always achieved: six agencies felt it had been successful, four said it wasn't, and one agency said it was too soon to evaluate success. Many of the comments note that there is not much revenue to be had from making the data accessible and/or that the data are not in a form that is much use to data users, and, presumably, unlikely to generate revenues for the user.

Over two-thirds of the eleven agencies that tried revenue sharing remain optimistic about its potential. Eight agencies said they would recommend revenue sharing to other agencies, with the main reason being to cover the costs associated with disseminating the data. Still, even some of those recommending revenue sharing hedge their bets and advise at least getting some non-monetary benefits, such as data for internal use or acknowledgment.

In contrast to the optimists, one of the eleven agencies that have tried revenue sharing recommended against it suggesting money makes agreements too complicated and suggested bartering for non-monetary benefits.

With one exception, the private sector was uniformly opposed to revenue sharing. The reasons for opposition range from the practical to the principled:

• Agencies are not accustomed to dealing with revenue-sharing arrangements and, if they are instituted, such arrangements are too hard to handle.
• The logic of revenue sharing is faulty, because there is not enough revenue to share.
• Because the public sector data is collected at taxpayer expense, should be available to all taxpayers, including firms. A private firm would be double paying for data generated with taxes.
• The private sector is serving a public mission by getting traveler information out to the public. The implication is that the private sector should not have to pay the public sector for doing the public sector's job.

The one firm that supports sharing a portion of its revenue with a public agency uses that practice in an arrangement that includes both exclusivity and the use of seed funding from the public sector. Another firm that has used revenue sharing in the past has abandoned the approach because it has not proved successful.

Exclusivity in data sharing

A traveler information business model based on exclusive access is used by about 15% of the surveyed agencies (PubQ9). Five agencies report that they have an exclusive arrangement with an organization for getting access to their data. Only one of the five agencies using an exclusive arrangement provides funds or subsidizes for disseminating data on an exclusive basis. The reasons for adopting the exclusivity model vary:

• Four out of the five agreed that it helped minimize the burden on the agency by creating a broker or distribution channel for the data.
• Two agencies felt an exclusive arrangement maximized the return to the agency for use of the data.
• One state DOT paid for development of traveler information displayed on kiosks in an exclusive arrangement with the kiosk owner. The DOT in turn gets reimbursed for its development effort with advertising revenue generated by the kiosks.
• One agency said it had an exclusive arrangement for convenience.

5.5 Lessons Learned about Data Sharing

The final stage of the interview provided an opportunity for respondents to reflect on lessons they had learned about the process of data sharing and to comment on other aspects of the topic.

Respondents from the public sector were asked to assess how effective their data sharing methods had been (PubQ19). Most agencies rate their performance in data sharing highly. Fifteen (of 33) give themselves a four on a five point scale and another six give themselves a top mark of five. Seven agencies rate themselves a three and five a two, but none put themselves in the lowest rating of one.

Despite their sense of satisfaction with their agencies' approach to data sharing, public sector respondents identified several lessons learned from their experience. The private sector respondents reflected on their experience, too. The principal comments of both groups are presented below and are segmented into operational, technical, and institutional topics. Of these, technical considerations appear to be the least of their concerns.

Operational Considerations

The need for marketing and outreach among agencies is needed to raise awareness about the potential for traveler information and experience in sharing of data.

• Several agency representatives said that agencies and companies need to be better informed about what data are collected.
• One private firm noted that agencies aren't always as cooperative as they need to be with regard to data. The respondent cited a dispute between a state DOT and the MPOs regarding dissemination of data that they share.
• One respondent said they desire to share more of what they have, while another said that many agencies do not seem to know about the benefits of sharing information.
• Agencies volunteered examples of the benefits of sharing. One instance of sharing of DOT video images with the transit agency had helped with a near-fatal incident, thereby increasing the transit agency's confidence in the DOT's data and video being shared. Because of the data sharing, another organization recently had an incident in one jurisdiction and the neighboring jurisdiction, knowing of the incident, coordinated their arterial traffic signal system to assist in smoother traffic flow.
• One respondent said that data is good public relations for an agency, and another said this is particularly true with visual data as opposed to numbers and maps.
• Agencies need to be more open with their peers on what works and what doesn't work, according to one private sector respondent. They aren't doing enough to share their problems or issues to help other implementers.

Handling revenue and expenses is always a major concern for both public and private entities, but no consensus exists on the recommended approaches.

• Several agencies said that the cost of collecting and making data available is high and some hoped that the private sector could share some of those costs. One of the firms was far from sanguine about assisting with public sector costs, stating flatly that the revenue sharing model doesn't work.
• Another respondent feared that revenue generation might be dampened by the consolidation of firms within the traveler information industry, leading to a monopoly situation.
• One agency respondent suggested that the public should provide data for free, a sentiment seconded by one of the private sector respondents. Another agency felt that competitive aspects of data provision should be put aside to do things for the common good.
• Several respondents noted the "infrastructure gap" [for collection of data], which may be in part related to the weak revenue potential for the information provided. Another suggested there needs to be more investment in outside trunk fiber connection.

Immaturity and volatility of the market for traveler information was reflected upon.

• The overall sense among agencies seems to be that the market is poorly developed with little revenue potential at the moment and, therefore, the public sector needs to plan realistically for bearing the cost.
• Several agencies mentioned that a characteristic of the immature market for travel data is its volatility. An example of that is the recent dramatic rise and fall in the web advertising market.
• Others mentioned that this volatility is also related to rapid technical change that can make plans obsolete very quickly. Change in market conditions and technology seems to be an issue for the nimbleness of government agencies and their relationships with the private sector.
• Another suggested that the DOTs should have their own data distribution channels because Internet service providers come and go, presumably interrupting data distribution and the development of a market.

Institutional Considerations

Cooperation among public agencies is not what it needs to be.

• One respondent suggested that there was a lack of leadership at the state level and another suggested that the state government needed to better understand what the public wants.
• On the other hand, a respondent from an eastern state noted that the state DOT had developed and provided software for its municipalities that worked well.
• Sharing of information on what others are doing around the country is desired, especially if such information can be tailored to relate to local conditions. One respondent wanted information about issues at the county level, which they said are different from those at the state level.
• Still, inclusiveness has its price. Every person you bring to the table adds exponential amount of time to process.

Policies and procedures facilitate data sharing both within and between organizations.

• Several find that it is a good idea to have agreements/policies with data users up front.
• A few respondents recommended one point of access for organizations using the data.
• One agency suggested that training is needed for people working in both public and private sector organizations.

Relationships between the public and private sectors continue to evolve.

• One respondent's agency is cautious about public-private partnerships. It would prefer not to have a private agency come and collect the data because of concern about the longevity of the firm.
• A private firm, on the other hand, argued that the public sector does not have to collect all information, and, indeed, the private sector doesn't count on just one data source.
• One firm spoke highly of its relationship with a state DOT, praising their dedicated technical people and the expertise they provided the firm in getting the data they needed.
• Another firm objects to exclusive arrangements between the public and private sectors with regard to access to right-of-way and to public data. In the eyes of another firm, such arrangements help a firm secure a competitive advantage [unfairly].
• Another firm recommends that the public sector keep clear about its mission in working with the private sector. In the respondent's experience, both sides are pretty reasonable, but that consultation is essential.

Technical Considerations

Data quality, quantity, and form of data are concerns voiced by some respondents.

• A common format and protocol along with data consistency and reliability are needed, according to one public agency respondent.
• There is a greater need for traveler information data to be reliable than for data for operational purposes because of the need to maintain user acceptance in the system, according to one agency respondent.
• The value of standards was reiterated by a couple of public sector respondents. One firm also cited the importance of standards, because they would help the firm provide a nationwide service for its customers.
• Nevertheless, one private firm thought data protocols and data dictionaries are useful for data exchange with agencies, but not essential, because the firm has its own software for extracting or converting data.
• Data that are more comprehensive both geographically and modally are needed. One agency respondent argued for greater coverage in small urban areas and rural areas because problems in those areas can be very extreme when they occur.
• One private firm noted that no public agency provides all the data needed, and that the private sector must augment the data, especially arterial and incident data. Another firm believes that the private sector will take whatever data the public sector has and improve upon it, and this should be acknowledged and enabled.

Creating traveler information from diverse information presents technical challenges.

• Most organizations are not familiar with XML, which could facilitate data sharing.
• Migration to open software is desirable.
• One agency felt that technical people need to get together to solve problems [more than they apparently do].
• Policies and agreements need to be flexible so that they can be modified if new technology becomes available.
• On respondent stressed the need for efficiency in data collection and sharing. The customized approach should be avoided and common formats should be used to ease integration.

6.0 POLICY ISSUES IN DATA SHARING

For traveler information to achieve its potential benefits for individual travelers and the transportation system as a whole, relevant data must be readily available for packaging and transmittal. Data generated by transportation agencies are the fundamental source for traveler information services provided by the public and private sectors. Thus, the sharing of public data is essential for traveler information services to flourish. A proponent of traveler information as a travel management tool, the U.S. DOT has an interest in their successful deployment. Consequently, understanding the dimensions of data sharing and policies that can foster the data sharing process are essential. This section discusses the philosophy and policy options for data sharing for traveler information.

6.1 Philosophies on Data Sharing

The term philosophy is appropriate because it stems from two premises about the role of government and the bureaucracy's responsibility for management of public resources. The philosophies of the actors in data sharing can be characterized as (1) the public view for "open" access, (2) the public view for exclusive access, and (3) the competitive private sector view.

Public Sector: The Open Door-Data Sharing as a Public Service

A majority of public agencies now take an "open access" approach to sharing of traffic and transportation data that they collect. Twenty-eight out of 34 agencies in the survey have no exclusive arrangements (PubQ9).

In their view information is collected, used and disseminated to achieve a variety of public purposes, such as improving safety, improving transportation system performance, reducing congestion, improving efficiency, saving energy, improving environmental quality, and improving productivity and economic vitality. Direct action is appropriate to get information to travelers if it is likely to help achieve the public purpose, thus VAR, HAR, and agency-owned or -operated Web sites are common. Thirty-two of the surveyed agencies (PubQ14) said they used one or more of these methods to get information to travelers. Some public agencies have set up phone services (often administered through contractors), a trend that may broaden due to 511 administration being placed in their charge.

With the philosophy of sharing data as a public service, the private sector can complement the public sector and help achieve public purposes. Public agencies generally must orient their services to the "general public," or, in other words, through broad distribution to a common audience. They are typically not as effective at reaching niche markets and do not have the resources or the tools to operate multiple delivery mechanisms both to broad and niche audiences.

Thus, agencies basing their data sharing on an open-door philosophy seek to enable and encourage as many pathways for information to reach the public as they can. It means that agencies collect and distribute information with their own resources as they deem necessary. They can also leverage the data by making them available to private firms and enlist their cooperation and support for achieving the public purpose. Public/private cooperation can also help the public agency avoid providing products and services in competition with private firms. If the private sector were willing to undertake the function of traveler information, the practice of good government would suggest that public resources not be used to do it. Furthermore, public entry into the market can dry up availability of private capital, thereby limiting the economic potential of the traveler information industry. On the other hand, if an agency finds the private sector to be unreliable in distributing the information that needs to be disseminated, it might need to do it itself.

Given the public purpose they seek to achieve, agencies using the open-door philosophy view appropriate partnerships with the private sector to be those that do not involve money flows either way and are done on a non-exclusionary basis. This:

• Avoids a burdensome procurement processes
• Allows for greater creativity
• Allows multiple information delivery mechanisms
• Allows competition for effective delivery to the public rather than focus on winning and retaining exclusive access to the information
• Often avoids confusion of roles and expectations

Therefore, the most effective partnerships will be those in which the public agency knows what it wants to achieve and negotiates with private firms so all are able to achieve their aims.

Of course, agencies still consider it good management to use agency resources wisely, and thus they will negotiate various forms of cost recovery (PubQ10) when they share data with other entities. These include having the agency pay for its own cost to connect, charging a fee for the data, charging for the agency's direct cost, or getting some in-kind contribution that is of value to the agency. Indeed, many agencies may experience advantages to having reciprocal data flows with multiple private partners regarding public cameras and private cameras, incident detection, and value-added data licensed back for public traffic management.

Public Sector: Data Sharing as a Business Proposition - The Exclusivity Model

A minority of agencies has chosen to sell or make their data available through exclusive or monopoly franchise arrangements. Only five out of 34 agencies surveyed (PubQ9) reported such an arrangement.

In contrast to agencies who feel obliged to share data with all who want them, the alternative philosophy views sharing data as a business proposition. In this view, the role of government is to wisely manage its resources and efficiently deliver services to the public. Agency data become a resource from which value can be extracted on behalf of taxpayers, and the most efficient means should be sought for the delivery of traveler information services to citizens. This view allows state and local governments to be free to establish whatever business arrangements they see fit. Conditions, needs and resource availability may vary significantly from one agency to the next, and these will be factored into the business arrangement. Furthermore, this view would hold that should the federal government place mandates or restrictions on local arrangements with the private sector, then federal resources should accompany them.

Practical considerations may lead some agencies to justify an exclusive arrangement. Some want to avoid dealing with too many private parties, and they would therefore prefer one point of contact, the most commonly cited reason in the survey. They may have invested in infrastructure but may not have budget available for information dissemination, or they may have no budget for either infrastructure or dissemination. They may want to encourage contribution to the public agency and its program. Reasons such as these can lead an agency to pursue procurement processes promising exclusive access to data and/or right-of-way in return for fees and/or promise of investment and delivery of services.

Private Perspective: Fairness through Competition

Firms in the traveler information business usually prefer to compete on features of their services over which they have control, such as timeliness, quality, usefulness, and user-friendly access to information. Exclusive access to data is viewed as anti-competitive because one firm is in a more favorable position for obtaining the raw material upon which traveler information services are based.

The seven firms surveyed were opposed to exclusivity, with the exception of one firm whose business model is predicated on access to public right-of-way utilizing public funding while exerting private ownership and control of the data collected. A variation that is condoned by the private sector is the competitive contract or franchise for operation of some portion of a traveler information system. In this case an agency uses a competitive bid process to outsource aspects of traveler information services that it chooses not to operate itself. Operation of a traveler information center (e.g. for the Metropolitan Transportation Commission in the San Francisco Bay Area) and the interactive voice response system (e.g. for Massachusetts Highway Authority in Boston) are examples. As long as firms are not excluded from obtaining the data on an equal basis from the contractor or franchisee, competition can be maintained. When exclusive access is awarded to a competitor, firms will try to find alternatives, such as negotiating with the competitor, collecting data with other means, or getting the public data in some other way, such as through another public agency. (Facts cannot be copyrighted, just the means and form of their presentation.)

In the competitive environment most private firms want a "level playing field" in arrangements with the public sector without one company getting special preference vs. others. In their view, information collected with public resources ought to be in the public domain. Taxpayers ought to be able to get access to it and the traveler information service providers are taxpayers. The exception should only be if there is a good public policy reason why the information should not be revealed, e.g., privacy, security, and safety. The public policy calculus regarding traffic information tilts toward open release and availability of the data, except perhaps personally identifiable information, "blood and guts," or information that could make conditions worse at an incident scene.

As a nascent industry, private traveler information providers are mindful of other public sector practices that can hurt their competitiveness beyond the issue of exclusivity. At the present time most are not making money outside the business model in which traveler information is part of a broadcast news or entertainment service funded through advertising. Consequently, having to pay for agency data or share revenues would dampen the prospects of these firms looking to expand beyond the current broadcast model. Thus, public agencies are finding that the value of their data to the private sector is lower than they initially expected. Factors that limit the market value of public data to private companies include:

• Private distribution is most likely part of a suite of services or information types.
• Private data always have to augment the public data (at additional cost).
• Public data does not have broad enough geographic coverage.
• Public data focus on freeways and not enough on arterials.
• Public data is "deep" to be able to control signals or meters; private sector wants "wide" - speed and conditions on multiple road and street categories in a wider area.
• Companies have to translate formats.
• The private firms perceive the greater value to be in packaging, marketing and distribution.

6.2 Policy Issues

The policy issues that have been identified that pertain to publicly gathered data to be used for traveler information include:

• Exclusivity of access to publicly collected information, e.g., franchise monopoly.


• Payment of fees for publicly collected data.


• Pay for hook-up to the public agency data.


• Sharing of private sector revenue generated from resale of the public data.


• Making the private sector's data with value added available back to the public agency for internal purposes.


• Use of "agent" or "broker" firms for distribution of data to other companies.


• Standards (data dictionaries, communication protocols).

The federal government has a number of approaches it can take regarding these issues. From benign to aggressive, it can:

• Do nothing


• Support information exchange, including reports, case studies, cross cutting studies, peer-to-peer visits and workshops


• Provide guidance regarding state of the practice through issuance of manuals and training courses


• Support development of consensus guidance through volunteer/peer committees of national associations


• Promulgate guidance for federally preferred practice


• Promulgate rules or propose legislation to Congress requiring certain actions as conditions for use of federal funds by transportation agencies.

Respondents to the survey expressed their opinions on the type of federal involvement that they would favor. Public sector representatives clearly stated their opposition to regulation or statute. As shown in Figure 8 the majority of agency respondents favor development of guidelines and information distribution or technical assistance (PubQ18). In general, private sector respondents shared these views with a couple of exceptions. One firm would like to see regulations that prohibited agencies from making exclusive arrangements for sharing their data with the private sector. Another firm was not interested in guidelines or regulations, feeling that both would slow down the development of traveler information services or meet resistance among the participants.

Both agencies and firms strongly support standardization of agency data and their transmission. However, the current status of the practice is not entirely clear. Whereas 20 of the 34 agency respondents claimed to have a common format for distribution of data to the private sector and other public agencies (PubQ16), that does not appear to be the norm for private sector respondents who receive data from agencies across the country (PvtQ11). Perhaps the difference can be explained in that many of the agency respondents are leaders in the traveler information field compared with colleagues in agencies not interviewed and have addressed the need for common formats.

Regardless of the current status, standards are favored by both entities. Twenty-six agencies believe there is a need for national or industry-wide standard format for travel-related data (PubQ17). The private sector agrees (PvtQ12), because firms often deal with multiple agencies that differ in their data formats. Some of their comments illustrate the point:

• "A standard data dictionary for common terms and usages and a standard digital output on how data are formatted are needed so that traveler information providers can connect regardless of the agency."
• "It would cut down on a lot of hardship and unnecessary cost for both public and private parties."

7.0 CONCLUSIONS

This section summarizes the results of the research on data sharing for traveler information and presents the principal conclusions that can be drawn from the findings.

7.1 Findings from the Research

Data sharing is addressed as both an abstract concept and a practical problem in the domain of traveler information.

As abstractions, the ATIS business models characterize public and private sector roles around the ownership and control of traveler information data. Agencies adopt the model that best suits their needs and resources, and, in the process, determine the extent of control they exercise over the use of publicly generated data. At the practical level public agencies are codifying their positions on control of public data used for traveler information in statements of policy and language in contracts with private sector recipients.

The National Weather Service has addressed the issues of sharing of public data with the private sector and comes out squarely in favor of an open and free model.

NWS minimizes the control it exercises on its data used by others and does not seek to profit in the dissemination. The economic benefits of the booming private sector weather information business are seen as validation of the NWS policy.

Many attempts at commercialization of public data in the U.S. and other countries have not fared well.

Information is inherently difficult to control and cannot easily be priced to reflect the true cost of collection. Examples of failures of the public sector trying to profit from its data are instructive to the field of traveler information.

Data sharing practices reflect dual objectives of transportation agencies.

Agencies share their data with the private sector and other agencies with two main objectives in mind: improving transportation operations through better interagency coordination and optimizing the use of the transportation system by providing information to travelers. In fact, enhancing interagency coordination in a region to improve travel conditions is the most important reason for sharing data. The surveys bear this out further in that nearly all agencies say that they share their data with other agencies.

Traveler information is a transportation management tool in which both the public and private sectors play a part.

By using agency-generated real-time and static data, both the public and private sector are active participants in providing traveler information even though their motives for doing so differ. Almost all the public agencies surveyed provide information directly to travelers, or are planning to do so in the near future, most commonly in the form of variable message signs, kiosks, highway advisory radio, and interactive voice response telephone.

Public data is generally not viewed as sufficient as a marketable commodity from which revenues can be generated. Businesses typically supplement the data received from a public agency with private data. Among the businesses receiving agency data, local television and cable TV stations, traffic reporting organizations, and local radio stations are the most frequent recipients. The most common types of traveler information provided by the private sector are traffic and road conditions, incident information, and planned construction information. Generally speaking, transit data are of less use in the traveler information business. Only about a third of the firms report transit delays, according to public agencies providing data to the firms.

Agencies protect their interests by placing conditions on access to data, but firms generally do not find these conditions to be onerous.

Making agency data widely available to other organizations does not mean that no strings are attached. Two or more conditions on access are common. Most frequently, an agency requires that it be acknowledged as the source, for example with display of an agency logo on a traffic video image. Many agencies restrict use in some way, to safeguard personal privacy or prohibit rebroadcast of graphic accident scenes for instance. Hardware and software requirements for receiving video or digital data are the third most common type of condition.

Technological limitations are barriers to data sharing for some agencies.

Sharing of agency video images seems to be the most problematic due to limitations to equipment at the agency. Indeed, problems with video sharing motivated one agency (VDOT) to develop a policy specifically aimed at limiting the number of private entities receiving a direct video connection at a VDOT traffic management center.

Formal policies and other tools are commonly used by transportation agencies to manage sharing of public data.

Over half the surveyed agencies have a policy in place now and several more have plans to issue one. The reasons behind the policies vary. The advantage of a formal policy for many agencies is that it provides a process for handling requests for their data. A number of others have policies aimed at facilitating the dissemination of traveler information.

In the absence of a formal policy, agencies are using other mechanisms to manage their data sharing activities. For example, some agencies report that they use contractual language to express their views, whereas others incorporate their views in training or procedure manuals.

Written agreements are often required when private firms acquire data from an agency. This is especially the case for video images and data in digital form. It is in these agreements that an agency would specify conditions, such as responsibility for costs and acknowledgement of the source of the data.

Sharing of data can entail costs to the public and private sectors, and agencies place clear responsibility on the receiving party to cover its own costs. Other cost-recovery mechanisms used are less common, but agencies frequently employ two or more mechanisms in data sharing relationships.

The general rule is that the private sector pays for its own hardware, software, and communication costs for accessing agency data. That mechanism is often used in conjunction with the private firm sharing its "value-added" information with the agency. Agencies have found other ways to recover their own costs or enhance their budgets. Although less common, direct reimbursement, in-kind contributions, and revenue sharing were reported by both the agency and private sector respondents.

Revenue sharing is a divisive practice among both public and private sector participants in data sharing.

While the public sector consists of optimists and pessimists on the potential for getting a share of revenues from firms receiving agency data, the private sector is opposed to the practice. While the idea of receiving revenues may appear attractive to some agencies, in practice it has not been much of a success. Private sector opposition is based on the practical difficulty of sharing revenues and the principle that public information should be available to all for free.

Exclusivity is an even greater touchstone among the private sector than revenue sharing.

The exclusivity model for traveler information is used by a minority of the surveyed agencies (~15%). Its main value is that it minimizes the burden on the agency by assigning responsibility for dissemination of data to a single firm. Private sector respondents implied or directly stated their opposition to exclusive arrangements. Their problem had to do primarily with the denial of access to data available to a competitor as opposed to "wholesale" or "broker" arrangements for subsequent distribution. The principles at stake are the right to access data collected with taxpayer money and the monopolistic and anti-competitive nature of franchises.

The experience of the agency and business representatives in the study represent lessons learned about the sharing of data in particular and traveler information in general from which other parties can benefit. Some of the lessons not previously addressed in this section include:

• More marketing and outreach among agencies is needed to raise awareness about the potential for traveler information and the experience of some of the trail blazers in addressing problems in sharing of public data.
• Immaturity and volatility in the market for traveler information market injects a note of realism into agency thinking. Relying on the private sector has its risks, but so does a totally agency-based approach to traveler information. Government agencies are frequently not nimble enough to address changing market and technology conditions.
• Agency leadership and interagency cooperation need to be improved. While that can be said, of course, for many public agency functions, comprehensive, multimodal traveler information relies on multiple sources. Leadership is needed for traveler information to be high on an agency's agenda and to bring other agencies to the table.
• Technical considerations do not appear to be the main issues in data sharing. Standards would be helpful, but are not perceived as major barriers. Of course, when it comes to filling "the data gap," lack of infrastructure for data collection is a problem that needs to be addressed. However, technical concerns appear to be solvable problems when it comes to simply sharing of existing public data.

BIBLIOGRAPHY

Booz Allen and Hamilton. 2000a. "Enhancing ITS Public/Private Partnerships in Wisconsin, Case Studies: San Antonio TransGuide MDI." Report prepared for the Wisconsin Department of Transportation.

Booz Allen and Hamilton. 2000b. "Enhancing ITS Public/Private Partnerships in Wisconsin, Case Studies: Seattle SmartTrek MDI." Report Prepared for Wisconsin Department of Transportation.

Booz Allen and Hamilton. 2000c. "Enhancing ITS Public/Private Partnerships in Wisconsin, Case Studies: Traffic Master - A Case Study." Report Prepared for Wisconsin Department of Transportation.

Booz Allen and Hamilton. 2000d. "Enhancing ITS Public/Private Partnerships in Wisconsin, Case Studies: TravInfo, The Bay Area Traveler Information Study." Report Prepared for Wisconsin Department of Transportation.

Booz Allen and Hamilton. 2000e. "Enhancing ITS Public/Private Partnerships in Wisconsin, Case Studies: The Japanese Vehicle Information and Communications System (VICS)." Report Prepared for Wisconsin Department of Transportation.

Boyd, S. 1999. "Minnesota Smart Work Zone Case Study." ITS American Proceedings. Washington, DC: ITS America.

Clinger, S. and D. R. Abedon. 2001. "Summary of Findings for CCTV Privacy Policy Survey and Examples of Existing Privacy Policy." U.S. Department of Transportation, Federal Highway Administration, Eastern Resource Center, Baltimore, MD.

Cross, G. 2000. "Travel Shenandoah: Lessons Learned in a Public/Private Partnership." ITS America Proceedings. Washington, DC: ITS America.

Hallenbeck, M.E., S. MacLachlan, and K. Godat. 1998. "Towards a Framework for Creating ATIS Business Plans." ITS America Proceedings. Washington, DC: ITS America.

Hallenbeck, M.E. 1998. "Choosing the Route to Traveler Information Systems Deployment: Decision Factors for Creating Public/Private Business Plans." ITS America, U.S. Department of Transportation Joint Program Office, Washington State Transportation Center.

Kosik, A. 1999. "TXDOT Information Sharing Policy." Data Archiving Workshop, College Station, TX: Texas Transportation Institute.

Kelley, J. nd. "Opportunities for 21st Century Meteorology: New Markets for Weather, Water, and Climate Information." Presentation.

Palmerlee, T. nd. "Public Agency Use of Proprietary Highway Network Geographic Base Files: A Workshop on Key Issues for Public Agencies." Personal communication.

Pretorius, P. and J. Markowitz. 2000. "ATIS Data Collection Guidelines Workshop, Problem Definition, Show Me the Data." Prepared for ATIS Data Collection Guidelines Workshop, February 9-10, 2000, Scottsdale, Arizona.

Schuman, R. et. al. 2001. "Scan of Existing Telephone Traveler Information Systems." 511 Policy Committee Retreat, Preparatory Materials. Washington, DC: AASHTO, APTA, ITS America, USDOT.

Stout, T., C. Poe, and M. Dixon. 1997. "Overview of the State-of-the-Practice in Disseminating Traffic Information by Advanced Traffic Management Systems," Research Report 1752-1, Texas Transportation Institute, College Station, Texas.

Weiss, P.N. 1997. "International Information Policy in Conflict: Open and Unrestricted Access versus Government Commercialization," in B. Kahin and C. Nesson (eds.), Borders in Cyberspace. Cambridge: MIT Press.

U.S. Department of Transportation, Joint Program Office. 2000. "What's Yours, Mine, and Ours: Overcoming Intellectual Property Rights Issues." Washington, DC.

To access an electronic version of this publication
and other ITS related publications visit the
ITS Electronic Document Library (EDL):
www.its.dot.gov

EDL Document Number: 13592

Visit our ITS Website
The ITS Joint Program Office:
www.its.dot.gov

Publication No.: FHWA-OP-02-001
HOIT/10-01

Footnotes:

[1] Responsibility for public sector interviewing was divided between Battelle, PATH, and Texas Transportation Institute. PATH and TTI were conducting a study of data sharing for four state departments of transportation (California, Minnesota, Texas, and Washington) similar in scope and timeframe as of the study commissioned by U.S. DOT. Therefore, a collaborative arrangement was established to share interviewing and survey results. Each study team produced separate reports.

[2] From the disclaimer of the National Weather Service which can be found on their Webpage at http://www.nws.noaa.gov/disclaimer.html.