What policies are useful in determining the cost of access to government information?

by / April 30, 1997
As agencies at all levels of government continue to move toward electronic creation and maintenance of virtually all records, one of the most intractable problems facing managers, whose records are subject to open records laws, is how much the agency may recover in costs for disclosing or affirmatively disseminating those records. For most access advocates, the answer is easy. Access laws are solidly based on a presumption that fees for access act as barriers to dissemination and, thus, run counter to the entire philosophy of open records laws.

In almost all cases, agencies should be allowed to charge only the marginal costs of dissemination. This means the direct costs associated with searching for and duplicating the records. In a paper environment, for example, a listing of all companies fined for violating certain workplace safety regulations might run to hundreds of pages, take an hour or two to gather, and take another hour to photocopy. That could run up charges in excess of $100 based on allowable direct costs. But if that same information is on a database, it may take a few minutes to find, and less than a dollar to copy off to a diskette. In other words, using a direct- cost model, electronic information can be much cheaper to the requester.

The paradox, however, is that it can also be much more valuable. And it is here that agencies frequently balk, asking why they should be giving away electronic data that may well be worth far in excess of direct costs. One good answer is that ours is a government of the people, by the people and for the people. Information amassed by government agencies is not the proprietary property of those agencies. It belongs, figuratively, to the people. If the people have already paid for its collection, analysis and maintenance, then why should they be required to pay again for access?

To the extent that individual requesters represent no group greater than their own self-interest, government can recoup the actual costs of giving them the information. But it may not recover the costs associated with creating or maintaining the information that has become subject to disclosure. Those costs have already been factored into the decision to collect and maintain the information in the first place. To go back to the earlier example -- the list of those fined for workplace infractions -- that list is probably not created solely to respond to access requests. Instead, it is kept as a record, or as pieces of information that can be assembled into that record, because the agency has some internal use for the information -- be it regulatory, administrative or purely statistical.

Perhaps the best defense of unfettered access to public records used
for commercial purposes has been given by Canada's Information Commissioner John Grace. He pointed out that companies that take public information, massage it and add value, create jobs in the economy. Those jobs, and the profits of those companies, generate taxes. And that is the trade-off. Information companies that enhance the value of government information and then resell it for a profit are certainly receiving no greater subsidy from government than companies who are able to use federal land at below-market value as a way of enhancing their profit margins.

The relationship between government and information brokers is similar to that between a supplier and a middleman who sees a way of taking part of that supply, enhancing its value and distributing it to a specific audience. From an access point of view, government should always consider universal availability of government information as its first goal. But the information broker may well be able to identify an information market that cannot be met by the government's distribution resources because the information must be massaged or refined to too great an extent.

However, by enhancing the government information, a broker may attract an audience that is substantially greater than the government could attract on its own. In other words, there may be a small subset of individuals interested in a specific type of government information, and direct distribution by mailing list, publications, or electronic dissemination may meet their needs. But by enhancing the information, by combining it with other information or other research or analytical capabilities, that same information core may attract a much wider audience which suddenly finds that enhanced information much more worthwhile.

This model assumes that the government information is hard to work with in its initial form -- that commercial enhancement is necessary to attract a viable marketplace. However, as database systems become more complex and layer upon layer of data is added at government's point of origin, the direct commercial marketability of the information may increase dramatically.

Geographic information systems (GIS) for cities, metropolitan areas and regions are great aggregations of information which in combination create analytical uses that make these databases potentially very valuable. The access model would conclude that such systems should be provided, to the extent that they do not contain any exempt information, at no more than the direct cost associated with making a copy for public disclosure. But common sense may dictate another answer.

While important analytical tools like GIS should not be kept away from the public, it may make sense to acknowledge their commercial value and design a cost scheme that helps account for that value. I have no trouble, as an access advocate, approving of a licensing arrangement or exclusive-use contract for a city's GIS when the end-user is a construction or architectural firm, a real-estate broker, or a consultant who advises companies on where to locate. Those kinds of commercial uses and any others of a similar nature could be required to pay something more than direct costs.

However, since government should not be in the business of selling information for commercial profits, perhaps the costs should be tied into a costing model that takes into account the cost of continuing maintenance and updating of the system. The cost of creating the system should not be a chargeable item since agencies create GIS and like systems for purposes other than public dissemination and the costs of creation have already been taken into account. Such projects should never be undertaken with the view that charging for public access will be the cash cow that supports the entire project from day one.

Where I would have a problem with this scheme would be if it were used to deny access for those who did not have a strong enough commercial incentive, or who just could not pay. GIS data could be invaluable to public- interest groups dealing with low-cost housing, health care, or the environment. But such groups are unlikely to have the funds to pay the higher commercially adjusted rates for such information. In such cases, government should subsidize the availability of the system; this could still mean collecting direct costs and providing a written agreement to restrict outside dissemination only to the public-interest group.

The simple explanation for access laws, ignoring the accountability trappings, is that government information belongs to the governed and should be made reasonably available to the public. But that simple and highly appropriate credo may need some refinement when information becomes more like a product and less like pure information.

Harry Hammitt is editor/publisher of Access Reports, a newsletter published in Lynchburg, Va., covering open government laws and information policy issues. E-mail: <>.

Harry Hammitt Contributing Writer