NYS ITS GIS Program Office
Geographic Information Systems Clearinghouse
Additional Information on the
NYS GIS Data Sharing Cooperative
A Supplement to The NYS GIS Data Sharing Cooperative document of 10/2/97
NYS Department of Transportation
October 2, 1997
Why not use the Federal Model?
This is a fair question, given that many states have chosen it as their model for sharing GIS data. In the federal model, data is released into the public domain with no restrictions, and either free or at no more than the cost of duplication. This model emphasizes maximum openness and sharing of data.
A significant down-side of the federal model is that there is no formal coordination of maintenance or custodial responsibilities for a given GIS dataset. The open aspect of the model encourages proliferation of datasets to many different users at minimum cost, but the user has no assurance of receiving the most up-to-date version, or more importantly, of receiving the corrections or updates made to the dataset by other users. This is particularly true of datasets downloadable from the Internet. The result is that many users will independently correct or update the dataset to meet their needs. This duplication of effort is very costly and wasteful. One need only consider the condition of the US Census Bureau's TIGER file and wonder how many different users, in both the public and private sectors, have spent time, energy, and money making independent and uncoordinated updates. TIGER data may be 'free', but what is the real cost of making it useable, and what has been spent, collectively, by everyone independently making it useable?
In theory, the federal model includes Area Integrators who endeavor to coordinate updates and corrections, on a voluntary basis, but evidence to date suggests that this is not effective. At a recent conference I learned that in North Carolina, a state often cited for their high level of GIS coordination, only 1/3 of the GIS data layers made available through the NC Center for Geographic Information and Analysis have a custodian committed to maintaining the data. This is perhaps not too surprising when one considers that under the federal model, the notion of data custodianship is voluntary, and many agencies who would be logical choices as data custodians may not wish to make a commitment, by themselves, on behalf of other users outside of their agency.
A second significant shortcoming of the federal model is that by being so 'open', it actually restricts participation. This paradox exists because participants must be willing to release their data into the public domain. Commercial data producers, utility companies, public/private partnerships, and many government agencies are typically unwilling to release their data under those circumstances. As a result, much of the best spatial data is not available through the federal model. These 'outsiders' from the federal model use marketplace mechanisms to control access to their data, so an interested user can expect to see a wide array of licenses, fee schedules, and restrictions, all used to allow access to the data without releasing it into the public domain. The Cooperative also uses some marketplace mechanisms, but in a uniform manner, to encourage public/private partnerships by protecting the interests of both partners.
The Cooperative can have broader participation because it uses a blend of the open sharing aspects of the federal model and licensing provisions of the marketplace models, presuming passage of the amended FOIL legislation. Open sharing occurs within the Cooperative, with costs never exceeding duplication fees. This follows the federal model. An improvement, however, is that data requests are always referred to the Primary Custodian of the dataset, so the requestor can be confident that they are receiving the most up-to-date and corrected version. A standard NYS GIS Cooperative DataSharing Agreement (license) is executed once by each member of the Cooperative, thus ensuring uniform 'rules' of data sharing and data maintenance for Cooperative members. The concept of Primary and Secondary Custodians ensures that clear and binding lines of responsibility exist for shared maintenance, revision, and redistribution of the data.
Shared maintenance is perhaps the most significant and cost-saving benefit of the Cooperative concept. The shared maintenance is not even a burden for Cooperative members; the Cooperative only requires that any error corrections or revisions that are performed on the data be shared back with the Primary Custodian. This mechanism harnesses the data maintenance actions that would happen anyway, to produce better data at lower composite cost for all Cooperative members. Primary Custodians will need to evaluate the contributed revisions and manage the overall updates of their datasets. In many cases, the contributed changes and revisions could serve as "triggers" to indicate areas in the database requiring a blanket update, such as from orthoimagery.
Partnerships with entities outside of the Cooperative are fostered. The ability of Cooperative members to license their data means that, unlike the federal model, members will be able to bargain in good faith as equals with outsiders for the return of value-added data to Cooperative members. Outsiders, either in the private sector or other non-members of the Cooperative, will have the confidence of receiving higher quality data from a Primary Custodian, while Cooperative members will be able to negotiate for value in return for the data, rather than a give-away. The fact that the data is not released into the public domain also removes a barrier that has prevented some private sector entities from partnering with the public sector on joint data development.
It's not about Revenue Generation
Since the pending FOIL amendment legislation includes the ability to charge up to the "fair market value" fees to commercial users of GIS data, some discussion on the use and intent of this provision is warranted. First, such fees are only applicable to commercial users. Within the Cooperative, fees are limited to costs of duplication. With the use of the Internet as a data delivery mechanism, it is expected that such costs will be near enough to zero to be ignored. Fees outside of the Cooperative to non-commercial entities are also limited to cost of duplication.
Second, commercial use fees are voluntary; in no case would an agency be compelled to charge any fees that it did not wish to charge. The fees are simply an option available to agencies. If such fees are used, the actual charge is determined by the agency, with the maximum amount being the "fair market value". There are provisions to ensure that unreasonable fees are not charged.
Third, the fees are not intended as a revenue stream for public sector entities to realize a profit or even to recoup their data development costs. Nearly ten years of data sales at DOT have shown that the magnitude of revenues to be made from GIS data sales are far too low to support a business model built on data revenues. The available evidence from other public sector agencies who charge fees for GIS data reveal similar results. Reasons for the low revenue returns are many, but include the marketing limitations inherent with public agencies, the small catalog of available data typical with most agencies, and a potential market consisting largely of other public agencies who resist paying data fees to another agency.
For all of these reasons, the Cooperative concept does not seek a model of revenue generation for its' members. This then raises the question; "What are the fees for, if not for revenue generation?" In a word, leverage. In many cases, a commercial firm wishing to acquire the dataset has something to offer in return. This would most likely be in the form of value-added to the database in terms of updates, enhancements, or extensions. The likelihood is high that such value-added features would be useful to the agency in particular, or the Cooperative as a whole. Licensing and fees provide the agency with real bargaining leverage to broker deals where the agency receives value-added features in lieu of fees. Without the license and fees, a commercial firm would have no reason to bargain away any of their value-added offerings. This bargaining ability gives public agencies a new tool to be better custodians of a taxpayer-supported database asset. This ultimately makes for improved efficiency and economy in the delivery of public service.
In situations where a commercial firm wants the data and has nothing to offer in return (say, an out-of-state credit reporting firm expanding their proprietary database), the fees, if imposed, serve to offset what might be viewed as a taxpayer-subsidized gift to a profit-making entity. The revenues from such fees are intended to offset costs of data improvement or enhanced access.
In summary, the fees to commercial users are voluntary, limited to "fair market value", and intended to be used only in the absence of other benefits that might be negotiated to the joint benefit of the public and private sectors.
Improved Data and New Data
One of the more difficult issues to sort out with respect to shared data maintenance is determining who 'owns' the revisions and corrections. The Cooperative concept handles this through a distinction between new data and improved data. These terms are more thoroughly defined and discussed in the Data Sharing Agreement, but a brief discussion here may be helpful to those unfamiliar with the concepts.
Improved Data are the corrections and revisions to existing categories of information, such as error corrections, improvement in positional accuracy, updating of existing categories, and other similar actions that do not create new categories of information. Improved Data are forwarded by all Secondary Custodians to the Primary Custodian, who may then evaluate the improved data and determine whether or not to incorporate it into the existing data. Since each dataset has a designated Primary Custodian, the Cooperative concept provides for that Primary Custodian to also be the Primary Custodian of the Improved Data. To do otherwise would result, over time, in a very complicated web of different owners of different pieces of improvements contributed to a particular dataset.
For New Data, that is, new categories of information which may be built on the foundation of existing data, the Primary Custodian of the New Data is the creating agency, unless some other agreement is reached to designate a different Primary Custodian. The Primary Custodian of existing data is not altered by the creation of New Data.
A Possible Hierarchy of Cooperatives
The Cooperative concept is scalable, that is, it could operate at several different levels. For example, a county might organize a county-level cooperative to deal with local government coordination of GIS datasets, while at the same time providing an entry point to the statewide Cooperative through a designated entity at the county level. Similarly, there could be sector-based cooperatives in vertical sectors, like transportation or real property, or horizontal sectors like metropolitan planning organizations (MPOs).
In order for multi-level cooperatives to function smoothly and without overlap, each lower-level cooperative would need a designated entry point to the state Cooperative. In this manner, a member of a local cooperative would access the state Cooperative through their designated entry point. This entry point would also coordinate the submissions of Improved Data back to Primary Custodians in the state Cooperative.
Large agencies who are members of the state Cooperative would conceptually operate in a similar fashion. It is typical of large agencies that there may be several different parts of the organization engaged in the use of GIS. In order for the agency to participate in a consistent manner in the Cooperative, some level of internal coordination within the agency will be necessary. The Cooperative requires the designation of a single point of contact at each member to ensure that the member is acting as an agency, rather than a collection of disjoint divisions, in honoring their obligations to the Cooperative.
The Cooperative relies on empowered custodians to work. There is no hierarchial organization to the Cooperative (other than oversight by the NYS GIS Coordinating Body), but rather a network of data custodians operating as equals. The terms and conditions of the Data Sharing Agreement reinforce the custodial responsibilities vested in Primary Custodians. Key characteristics of Primary Custodians include:
- Autonomy to decide how and whether or not to incorporate improved data received from Secondary Custodians. An agency is Primary Custodian of a dataset by virtue of having created it, and as such is presumed to have a business need to keep it maintained. They can use their own judgement on determining the best way to do this.
- Sole source for obtaining a particular dataset. Since each dataset will have a single Primary Custodian, and since Secondary Custodians are not permitted to redistribute data obtained from Primary Custodians, there will be no ambiguity on where to obtain a dataset. Likewise, there will be no ambiguity as to whether the most recent version is being obtained.
- Ability to opt for public domain distribution. The Cooperative does not require Primary Custodians to limit distribution of their datasets to the Cooperative. Members who desire to release their own data into the public domain are free to do so. In doing so, however, they will not be able to take advantage of many of the benefits of the Cooperative, since data in the public domain is available to all with no restrictions or obligations.
- Bargaining leverage to negotiate with outsiders for value-added improvements to the data. This enables members to obtain real value in exchange for their data, potentially lowering their costs for improvements that they might otherwise need to purchase or do without.
In summary, the Cooperative does not prescribe to Primary Custodians what they must do with their own data (other than making it available to other Coop Members and adhering to standards), but rather tells them what obligations and limitations they have with other members' data.
The Bottom Line
The New York State GIS Data Sharing Cooperative concept is a bold new idea. To the best of our knowledge it has not been tried anywhere before. It is designed to overcome the barriers that have prevented many GIS players from actively participating in a shared GIS environment, while at the same time having enough flexibility to adapt to a very dynamic and fast-changing technology. The Cooperative concept allows us to leverage the ongoing actions of many GIS users to create a stronger data framework.
Let's make it work.