Government Information in Canada/Information gouvernementale au
Canada, Number/Numéro 18 (August 1999)
Threats to Public Access to Federal Government Publications in Canada and
the United States
by Katherine Prophet
The Canadian Experience
Economic Theory and Access to Government Publications: Cost Recovery, Privatization, and Commercialization
The trends of cost recovery, privatization, and commercialization are intertwined: they all arise from the need for government agencies to achieve fiscal balance. In Canada, with the drive to lower costs and increase revenues, government departments can privatize their information, usually by licensing the private sector to use government data to create a "value-added product," or departments can charge for it themselves, commercializing information to recover costs.
Considering the increases in user access fees imposed by government departments, Bruce Morton observed:
Overall [during the period of 1984 through 1993], government departments and agencies were directed to make more information accessible while at the same time being told to generate more non-tax revenues from it. Therefore, the impulse to provide the public with low-cost information was offset by an even stronger impulse to generate income from government information. (17)
Kirsti Nilsen sought to link federal government policies of restraint and cost-recovery espoused in the mid-1980s to the steep rise in prices imposed by Statistics Canada for its print publications and for access to its electronic databases. She maintains that "While these policies were not labelled as 'information policy,' they were, in fact de facto information policies which were applied by government agencies to their information products." (18)
Nilsen's study shows a positive correlation between the 1986 Budget Speech and sharply increased prices for Statistics Canada print publications. This precipitous price hike occurred while more of its data was being offered only in electronic form at even higher prices. (19) She concluded that the correlation was strong enough to indicate the Statistics Canada move to higher prices and expensive electronic data dissemination was a direct result of government cost-recovery initiatives. (20) Nilsen warns, "Governments which philosophically espouse market solutions to achieve fiscal balance will formulate information policies which identify information as a resource and a commodity. Pricing which generates revenue is encouraged." (21)
Monty identifies Statistics Canada as the first agency seeking to recoup costs by raising prices. She points out that, under the Statistics Act, the agency can compel a private citizen to supply data, but that this data is now used to recover costs or even make a profit. (22) Discussing the Canadian census, Monty points out: "Most of these data are now in electronic format and it costs thousands of dollars to obtain access to it. The cost of purchase is beyond even large institutions, let alone depository libraries or individuals." (23) Morton reports that in 1987, in its release of the 1986 census, Statistics Canada offered 60 percent fewer computer-readable products than it had for the previous census, while charging 1,500 to 9,000 percent more. (24)
During this period of fiscal restraint, the survey and mapping functions of the Department of Energy, Mines, Resources were privatized, as were the specialized weather services. (25) By 1991, the Department of Consumer and Corporate Affairs was distributing its database via private vendors through licensing agreements. (26)
The Treasury Board Government Communications Policy published in December of 1996 reflects the impact of cost-recovery policies on the production and dissemination of government information. In its policy statement, the communication document affirms, "The responsibility to provide information is inseparable from the nature of representative government. Adequate information is essential in order that the public--individually or through representative groups or members of Parliament--may understand, respond to and influence the development and implementation of government policies and programs." (27) It continues:
Good communications is fundamental to the achievement of government objectives. Communications, planning, coordination and execution are an integral part of the management process of government. Therefore, it is the policy of the government:
The communications policy does uphold the democratic ideal of access to government information in its policy statement, but its policy objective is different in tone: "The objective of this policy is to ensure the effective management of government communications." (29) "Effective management" apparently incorporates user fees under section 4, "Availability and dissemination of information." The principles expressed in this section affirm the initial policy statement and cite the Access to Information Act as the "legal framework for openness in government by giving Canadians the right of access, subject to limited and specific exemptions, to government information." (30) Then, the document introduces the idea of user fees:
However, the provision of information is costly and should be undertaken only where there is a clear duty to inform the public or where the user is willing to pay for it. The full cost of providing information to serve the proprietary interests of individuals should not be borne by taxpayers at large.
In assessing the cost of making information available for purchase by the public, institutions should take into consideration the full costs of collecting, compiling, preparing, producing and disseminating information. (31)
The concept of cost recovery through user fees is consequently part of the policy.
The directive points out that the Access to Information Act does not "legally apply to published material or material available for purchase." (32) The implications of this sentence are quite far-reaching. Nilsen reported that government agencies were informed in 1988 at a government database colloquium "that when a government database is marketed commercially it is considered to be 'published.'" (33) This means requests can therefore be refused, and that those wanting the information must then pay a vendor charge that is higher than the fees called for by the AIA. With this interpretation of the act, the originating agency not only benefits from vendor licensing fees, but also circumvents the costs associated with access requests. (34) The maze of contradictory government information directives, alluded to by Morton, is quite apparent in the Government Communications Policy.
Section 4.5 deals with publishing. It states that copublishing should be arranged with the private sector "where practical." (35) If private copublishing proves to be impractical, CCG-P (Canada Communication Group-Publishing) "may [emphasis mine] arrange to publish the work directly." (36) This section further stipulates that "for purposes of the Depository Services Program, institutions are required to provide CCG-P with copies of each item that they publish." (37) Because this document is a statement of policy by a federal department, and is not a law, any directives given do not have the force of law. Institutions are "required" by the policy, but not by a law which entails consequences for non-compliance.
The Treasury Board Cost Recovery and Charging Policy delineates guidelines for cost recovery and charging decisions by departments and agencies for a broad spectrum of goods and services, including food inspection, computer services, and information. At the core of the guidelines is the philosophy that "there must be a relationship between the fee charged and the cost of the good or service, or the value of the privilege provided to clients. Charges help determine the proper scale of delivery by applying a market test of underlying demand, thus tending to eliminate the over-use that often exists with 'free' services." (38)
The policy statement reflects this emphasis: "It is government policy to implement user charges for services that provide identifiable recipients with direct benefits beyond those received by the general public..." (39) The Cost Recovery and Charging Policy aims to efficiently allocate resources. Reflecting government fiscal policies, it seeks
to promote an equitable approach to financing government programs...by fairly charging clients or beneficiaries who benefit from services beyond those enjoyed by the general public. This may allow a greater share of general tax dollars to be devoted to activities that benefit the general taxpayer, or to reduce the debt [emphasis mine]. It may also facilitate improvements in the delivery of specific cost-recovered services. (40)
The policy states that not all government activities should entail user charges. Programs that function for the "general benefit of all" (41) or are intended to assist recipients should be supported by taxes. The cost recovery policy cites a background paper, User Charging in the Federal Government, prepared by economists Richard M. Bird and Thomas Tsiopoulos. This paper formulates the criteria used by government agencies to assess when user fees are appropriate and whether a government program or activity constitutes a public or private good. (42)
The paper has two purposes: to explain the economic and managerial rationale for user charges and to provide guidelines for determining where charges should be levied and what those charges should be. (43) The authors cite the economic history of the 1970s and 1980s and indicate that there should be a continuing policy of cost recovery.
Until the mid-1970s, the growth in expenditure at the federal level was financed relatively painlessly by an increasing stream of tax revenues produced by economic expansion. As economic growth slowed, expenditure growth was financed partly by deficits. More recently, federal expenditure growth has not only halted but has been reversed. In the current economic and fiscal environment there is little chance that any significant new expenditures can be financed out of general tax revenues. Indeed, there is, and will likely continue to be, increasing pressure to reduce not only deficits and expenditures but also taxation. In these circumstances, it is important to carefully examine the rationale and structure of all federal expenditures and revenues, including user charge policy. (44)
The basic principle underlying the paper is, as the authors note, the precise opposite of the view that charging for public services is the equivalent of a "revenue grab." Bird and Tsiopoulos unequivocally state that "whenever possible and desirable, public services should be charged for rather than given away." (45) User charges, the authors stress, promote economic efficiency and improve accountability.
In determining which services should incur charges, User Charging notes that "to some extent, the continuum between purely "public" goods and purely "private" goods is matched by a similar one between general fund financing by taxes on one hand and user charges on the other." (46) Bird and Tsiopoulos use six different characteristics to determine an activity's location on this continuum: rivalness, excludability, economies of scale, lumpiness, externalities, and social objects. These are economic terms that librarians need to comprehend.
It is important to understand these criteria because they are being used to assess all federal activities and goods, including the production and distribution of government information. As Harry Hillman Chartrand writes, "According to Treasury Board's experts [Bird and Tsiopoulos], user charges are appropriate for only some federal goods and services. They propose six criteria to determine whether or not such charges are appropriate rather than relying on general taxation. ...They may prove to be a rusty nail on the up-ramp of Canada's information superhighway." (49) Chartrand uses the six criteria to argue that federal information should not be subject to charges.
Chartrand maintains that information is non-rival in consumption, that in reality it is "collaborative or additive in consumption." (50) He points out that in research, one researcher's information leads to more information produced by other researchers, generating more consumption and even more production. He continues that information is implicitly non-excludable and argues that if it were truly excludable, there would be no need for copyright. To strengthen this argument, he states that "even copyright, however, does not protect information as such but rather its fixation in material form, i.e. ideas are not protected but rather the specific form of their expression." (51)
Chartrand suggests that the "additive" nature of information projects the expectation of economies of scale in its production. He uses the example of huge databases like the taxation and health statistics, the Census, and environmental data. Once created, these databases can then generate, at a much lower cost, new information through electronic cross-referencing. For instance, Census data could be correlated with environmental data to uncover any "meaningful relationships between life expectancy and selected environmental quality indicators. This, in turn, may direct public policy towards limiting the health cost impact of negative environmental factors..." (52)
The federal government has made an enormous initial investment in developing its data systems, and this investment must be maintained with updating and staff training. The size of this initial federal investment demonstrates lumpiness. (53) Chartrand links this lumpiness in the costs of producing federal information with the identification of a public good. He uses the example of the Census, indicating that it is unlikely that any private sector body could afford to conduct a Canadian census. He also points out that the private sector would not have the coercive power of the Statistics Act to elicit a meaningful number of responses.
Considering the external effects of information, Chartrand turns to economic theory which assumes "that for a perfectly competitive marketplace to exist there must be 'perfect information.' If buyers and sellers operate in ignorance or with unequal access to relevant information then market outcomes will be distorted." (54)
According to Chartrand, securities exchange laws worldwide recognize this concept by making insider trading illegal. He continues by arguing that federal information often generates specific external benefits including "'serendipity' in scientific and social research, enhanced business efficiency, and education." (55)
Chartrand avers that scientific and social research manifest a high degree of serendipity, embodied in those unexpected flashes of realization when heretofore unrelated data and/or observations coalesce in discovery. Chartrand asserts:
If federal information is available at little or no cost, the researcher can afford to follow a "hunch" and/or by accident come across such a correlation. If federal data is costly, the researcher may not be able to afford to play. (56)
The Humanities and Social Science Federation of Canada (HSSFC), the Canadian Association of Research Libraries (CARL), the Canadian Association of Small University Libraries (CASUL), and the Canadian Association of Public Data Users (CAPDU) struck an agreement, the Data Liberation Initiative, with Statistics Canada to address, at least partly, the problems posed to research by user fees. For a greatly reduced fee, Canadian academic institutions can access Statistics Canada electronic data collections; individuals in these institutions pay nothing. Even though the costs are less, the institutions still pay fees. As Alan MacDonald, who views the initiative in a positive light, reports, some fear an undesirable precedent has been set. (57)
In the business arena, information is essential: the more information a business has, the better its decisions. For Canada to be competitive in the global market, its information costs must be competitive. Chartrand points out that "...Crown copyright and more limited exemptions from copyright infringement appear to make Canada less competitive relative to the United States. Federal information user charges would tend to increase this burden." (58)
Chartrand extends his case for the external benefits of information by citing the public library system. He states that society as a whole benefits from a more literate population and that public libraries therefore should be deemed a merit good because they promote literacy and an informed populace.
To the degree federal information user charges limit access to information, they limit the educational benefits of federal information restricting them to those with the ability to pay. In welfare economic terms, this represents a "regressive" outcome, i.e. those needing the educational benefits of federal information most, that is the poor, are least able to afford access to such information. (59)
By the standards of the final criterion, socio-political objectives, access to government information engenders an "information democracy" (60) in which citizens can make informed electoral decisions and government can be held accountable for its actions. User charges inhibit this scrutiny and limit citizen ability to judge government decisions. Another positive socio-political objective fostered by eliminating or minimizing federal user charges would be the development of the Canadian information infrastructure. Erecting barriers to access with charges jeopardizes Canada's profile on the Net and impedes the growth of a "Canadian information sovereignty." (61)
Using the Treasury Board's criteria, Chartrand has aptly demonstrated that federal information is a public, rather than private good, and along the continuum envisioned by Bird and Tsiopolous, it falls more decidedly under general fund financing than user charges.
Bird and Tsiopoulos also touch on privatization in their background paper. In proposing the move "to a more market-oriented approach," (62) they state:
If the public sector is providing a service that could be equally well provided by the private sector, the activity in question should probably be privatized. This conclusion may be valid even when supporting public policies are needed. (63)
They conclude this discussion by observing, "More broadly, however, user charges and privatization are best seen as just two of many approaches to the same policy objective--getting the most for Canadians out of the resources available." (64)
Government publications librarians should understand these criteria so they can effectively argue for public access and a strong depository library system in the face of economics and its terminology. They should be ready to prove that government information is a public good that should be fairly and easily available to the public.
Departmental Autonomy and Decentralization
Compounding the threat posed by cost recovery and privatization is the trend toward departmental autonomy and decentralization. Through the Depository Services Program, depository libraries in both countries have been the conduit by which members of the public access government information, but this system is now faced with a confusing labyrinth of independent departments, whose names can mutate and whose charges can fluctuate. Vivienne Monty writes:
Departmental autonomy is yet another issue for concern. Proponents of departmental autonomy believe decentralization will make government more efficient. As a consequence of this autonomy, agencies do not have to publish through the CCG-P; they can go to private sources for cheaper production. Internally, however, the payment they get for publications from DSP is at reduced cost. The departments consequently say that giving materials to the DSP actually costs them even if they are reimbursed to some degree.
Departmental decentralization has reinforced cost recovery activities and ultimately increases the price of information. (65)
She also points out that this decentralization has exacerbated poor bibliographic control; therefore, materials are more difficult to identify and obtain. She states that "many fear a return to the bibliographic chaos that existed before the establishment of the CCG-P." (66)
Sharp Curves on the Information Highway
The economic and organizational hurdles encountered by those seeking government information, as well as by the libraries that serve them, are further complicated by the new technologies. Nilsen observes that "technology enhances bureaucratic control and encourages the imposition of higher prices rather than increased access." (67) For example, in 1995 McMahon asserted that Statistics Canada used electronic dissemination to raise money. He cited the Census profile series which, for each province, was priced at $125 per print copy, versus $250 for the same publication in electronic format. (68)
The EPP document acknowledges in its background statement that "increasing demands for information coupled with financial cutbacks have forced government publishing programs into transition. As a result, many government departments are changing their modes of dissemination from the print to electronic medium and, consequently, planning and developing new electronic publishing infrastructures." (69) The implementation of the new technologies needed to keep up with this change poses other serious problems for the depository libraries as they attempt to keep access flowing.
Two recent studies examining these issues identified areas of librarian concern about Internet access. A study conducted by Liwen Qiu Vaughn and Elizabeth Dolan surveyed all full and selective depository libraries in the DSP, while the Electronic Publications Pilot (EPP), a joint initiative of Statistics Canada and the DSP, encompassed selected libraries. The 1996/1997 Dolan-Vaughn study aimed to assess the preparedness of libraries for the move to electronic formats, while the 1996/1997 EPP sought to gather information regarding the adjustment of library staff and clients to electronic environments like the Internet.
Dolan and Vaughn quote Gary Cornwell, who, when addressing the U.S. situation, pointed out that "while electronic publishing and dissemination will greatly reduce costs for information producing agencies, it will place an undue burden on libraries and will unquestionably impact public access to federal government information." (70) Their survey confirms his observations, particularly for Canadian public depository libraries, which constitute the most common type (at 51.5%) of federal depository library. They conclude that "...the public libraries, which serve larger populations and have more direct contact with the general public, are disadvantaged in technology--the prerequisite for electronic access to government information." (71)
According to Dolan and Vaughn, computer facilities in public libraries lag far behind what is available in academic and government libraries. Academic and government libraries had ten times more computers per 10,000 people served than did public libraries: 20 and 17 PCs, respectively, to 1.5. The public libraries also had fewer computers with greater than 8MB of RAM. Ample RAM (active memory) is necessary to efficiently run Internet applications. The average public library had only one PC with Internet access for every 20,000 people it served, while academic and government facilities boasted 16.4 and 3.2, respectively. In general, the authors note, public libraries had lower bandwidth connections.
Public libraries also had fewer and slower printers, the most common type of printer (41%) being the dot matrix. Dolan and Vaughn report "...there is only one printer for every 20,000 people served in a typical public library while there are 8 and 12, respectively, in academic and government libraries for the same number of people served. The difference is highly significant." (72)
The survey revealed that many users often printed out electronic documents because print documents are easier to read and because many libraries impose time limits on the use of Internet workstations. Even though Internet access was free in 89.2% of libraries, printing was not, with 67.2% charging fees for the service. Eighty-eight percent of public libraries charged users for printing. As Dolan and Vaughn note, "Many survey respondents acknowledged the potential of the Internet for timely access but expressed concerns regarding the transfer of printing cost from government to libraries and/or users." (73)
Summing up their findings, the authors asserted:
Results of the study demonstrate that in accessing electronic government information public library users are (or will be) significantly disadvantaged compared with their counterparts in academic and government libraries: fewer and lower quality computers and printers are available to them, narrower bandwidth Internet access is the rule, and they are more likely to have to pay for services. ...A threat to equitable access to government information is imminent if the government moves too quickly from print to electronic form to deliver and disseminate its formation [sic]. (74)
Dolan and Vaughn conclude: "Both positive and negative effects of an electronic DSP should be investigated before full implementation takes place." (75)
In their DSP report, entitled Electronic Access to Canadian Federal Government Information and based on the same 1996/1997 survey, Dolan and Vaughn add that print formats were used more frequently than electronic. Reasons for the relatively low use of electronic documents were extrapolated from respondents' observations and included a lack of computer equipment coupled with a shortage of staff time to assist users. Survey results also indicated that many depositories were short of the trained personnel required for user assistance.
The Electronic Publications Pilot, a study conducted in 1996-1997 to test and assess access to Statistics Canada's website did not attempt to survey all depository libraries. Participating librarians were selected from libraries with heavy usage records and more hits on the Statistics Canada site. The report admits that "the results of the EPP are not necessarily representative of the whole DSP community. Not only were libraries self-selected, but they were required to meet certain criteria as a condition of participating in the EPP. Thus, it is possible that the sample libraries were more prepared for electronic dissemination than the DSP libraries in general." (76) In spite of this bias, the Electronic Publications Pilot still unearthed feedback and information pinpointing problems with publications dissemination via the Internet.
Just getting connected to participate in the pilot proved problematic for many libraries. Over half of the libraries (57% on the first questionnaire, 50% on the second) experienced difficulty connecting to the EPP sites, with 41% of those libraries having connection problems "frequently" or "very frequently." (77) About two-thirds of the participating public libraries had connection problems. This finding complements Dolan and Vaughn's cautions about public library capabilities and preparedness. The EPP report notes that the academic and government libraries tended to have direct Internet connections, while the public libraries were more likely to use Internet Service Providers.
The EPP concludes that the major concerns expressed by librarians relate to resources and training, the Statistics Canada information web sites, publications selection, and access. (78) Resource barriers included deficiencies and shortfalls in hardware, software, finances, and human staffing. Eighty-one percent of EPP respondents continued to offer alternatives to online access, like publications stored on the hard drive or shelved print copies. The report attributes this to a lack of human resources. It notes that regardless of the number of computers with Internet connections, the "proportion of libraries that offered alternatives was high." (79) The document reports that one-fifth of the libraries provided offline options because of a shortage of available staff to assist users. When asked about any modifications necessary to meet the demands of the Web format, 50% of responding libraries stated that "they needed more time for electronic dissemination" (80) in order to facilitate its effective use. Forty-seven percent called for more staff.
Staff training issues went hand-in-hand with staffing shortages as a major human resource issue. While 80% of respondents judged training to be necessary for the transition to electronic dissemination, only 19% of responding librarians wrote that they had conducted some training. Many cited the need for training with Adobe Acrobat (85%) and with the Internet (43%). The problems surrounding the training issue are delineated in the report:
It is clear from the EPP that librarians, who mediate access to government publications, must have the appropriate skill sets for electronic dissemination. In addition, they must be able to train end-users in the electronic medium. However, we heard from librarians that, with budget cutbacks, they had less time to learn to work with electronically disseminated publications. At the same time, 58% of librarians stated that the electronic delivery of publications had increased the time they spent teaching and mediating the use of electronic products, compared to print publications. ... Currently, the lack of training is a substantial barrier to maintaining access to government publications in electronic format. (81)
The lack of hardware resources was apparent with 69% of libraries expressing the need for additional Internet-connected PCs. Most libraries, 85%, also required software upgrades. Compounding the problems posed by the pressures to acquire more and improved resources was the lack of money. Libraries reported that their financial resources were diminishing. When the fact that this pilot involved self-selected libraries is taken into account, these findings constitute more than a mere pothole in the road to efficient and equitable electronic access.
Beyond their own resource problems, the libraries expressed concerns about supply-side issues. The study sought to ascertain which publications would be suitable candidates for dissemination in electronic format only. Based on respondent feedback, the EPP indicates that length and demand should be deciding factors. The higher the demand or the longer the document, the less desirable it would be to offer it only over the Web. If there were high demand for a particular publication, like Juristat, in a library with few Internet connections, access to that publication would be consequently reduced. (82) If print were available as an alternative, this lessened access could be alleviated. Longer documents take longer to download and print, and many patrons want print copies. This renders the more lengthy publications also less suited to electronic dissemination.
The EPP pinpoints issues of archiving and permanent access:
There was overwhelming support from EPP participants for a centralized electronic archive. Librarians were also very clear in their wish for permanent electronic access to this archive. The main concern that was raised with respect to electronic dissemination and archiving was the ability to access and read publications in the future given rapid technological change. One hundred year old print products are accessible and readable, but will that be true for electronic publications in the future? Librarians expressed their concern that these issues be tackled immediately as they may dictate decisions such as the format in which publications are currently produced. (83)
Such concerns are well-founded. In the United States, National Media Lab tests demonstrated that the life expectancy of CD-ROMS can fall anywhere between five and one hundred years, (84) with average-quality CD-ROMs deteriorating in five years. (85) Doculabs put eight types of CD-ROM media through an accelerated aging process. After aging, three of the eight were completely unreadable. (86) Magnetic tape has a life expectancy of only ten to thirty years.
As Laura Tangley points out in U.S. News & World Report, even if the storage media display optimum longevity, the hardware and software required to read them in the future might be found only in museums, if at all. She recounts that Donald Waters, director of the Digital Library Federation, terms the problem a "time bomb." (87) The bomb has already gone off in some places. At the New York State Archives, Margaret Hedstrom tried unsuccessfully to read land use information on indecipherable magnetic tapes from the 1960s. Satellite photos of the Brazilian Amazon, taken in the 1970s, are locked in unreadable magnetic tapes. (88) These photos would be invaluable in tracking deforestation trends. (89)
Librarians' fears about the permanence of archived electronic data are proving to be true. In addition to a centralized archive, librarians expressed that they would prefer one centralized site for Web access.
The EPP outlines some final concerns and considerations like the pivotal question of who will pay for providing electronic access. Because many public libraries must pay a commercial Internet provider for access, over a fifth (21%) charge their clients for Internet use. (90) Because many libraries still use dot matrix printers, printer upgrades must be made. "If libraries are expected to upgrade their printing equipment in order to provide patrons with the same quality of publication they have come to expect, many will begin passing on the costs of this service to the patron." (91)
The Dolan-Vaughn and EPP studies show how many crucial problems need to be addressed in implementing Internet access. They both point to the necessity of proceeding cautiously and systematically into the establishment of an electronic Web-based approach so as not to jeopardize further the already eroding access to government information.
Next: Proposal for a Revised Model Depository System