Keywords
Citation
Sloan, B. (2000), "Cost sharing", The Bottom Line, Vol. 13 No. 3. https://doi.org/10.1108/bl.2000.17013caf.002
Publisher
:Emerald Group Publishing Limited
Copyright © 2000, MCB UP Limited
Cost sharing
Keywords Consortia, Libraries, Costs, Cost allocation, Public libraries, Academic libraries
Previously, I introduced the topic of consortial cost sharing by illustrating how the membership composition of a consortium can complicate cost-sharing models, using the Illinois Library Computer Systems Organization (ILCSO) as an example. I outlined how ILCSO has apportioned costs among its membership since Fiscal Year 1988, and concluded the column by summarizing three key lessons that might be garnered from the ILCSO experience to date.
How do consortia other than ILCSO apportion costs? Over the past few years, I have gathered information on this topic, to aid the ILCSO Assessment Task Force in the process of calculating annual ILCSO membership assessments. Following are brief summaries of the methods that 21 library consortia use to apportion costs among members.
Black Gold Library System
The Black Gold Library System (Ventura, CA) divides the costs of their shared Dynix system using three factors: patron count, circulation, and number of ports.
Committee on Institutional Cooperation
The CIC's Program for Library Initiatives divides operational costs equally among members. Additional project costs are divided one of two ways: equally among members, or based on Full Time Equivalent (FTE).
Colorado Alliance for Research Libraries
CARL calculates membership fees and assessments using three factors: a flat membership fee ($25,000 per institution), plus a proration of expenses based on relative percentages of ports and items held.
Council of Prairie and Pacific University Libraries
COPPUL is a consortium of university libraries in Western Canada. COPPUL has a fixed structure of membership fees based on the overall operating budget of the entire university. There are nine levels of fees, starting with a $1,500 fee for a university with an operating budget of $25 million and under, up to a $10,500 membership fee for universities with an operating budget in excess of $350 million. COPPUL had previously based the fees on library materials budget, but did not find that method satisfactory.
Electronic Library Network
The Electronic Library Network, a consortium of academic libraries in British Columbia, uses two factors in its cost-sharing model. They calculate each library's relative percentage of total acquisitions budget for all members, and each library's relative percentage of total FTE enrollment for all members. The two factors are weighted, so that the relative percentage of acquisitions budget counts for 40 per cent, and the relative percentage of FTE is weighted at 60 per cent.
Five Colleges Consortium
The consortium (consisting of Amherst, Mt Holyoke, Smith and Hampshire Colleges, and the University of Massachusetts at Amherst) uses several methods. They divide the costs of their ILL shuttle service evenly among all participants. Shared database subscription costs are prorated based on usage. The assessments for running their shared online system are based on, vaguely, the comparative size of the institution.
Florida Center for Library Automation
While the bulk of FCLA's operating costs are funded through a state appropriation, the consortium does divide the costs of reference databases on a prorated basis. The license fees for these databases are prorated based on each library's percentage of the total library materials budget for the consortium as a whole.
LEIAN/NMCAL
The Llano Estacado Information Access Network (LEIAN) and the New Mexico Council of Academic Libraries (NMCAL) have teamed together for a group subscription to FirstSearch. They divide all costs based on FTE enrollment.
Montgomery County Library and Information Network Consortium
MCLINC is an eight-member public library consortium in southeastern Pennsylvania. The total costs of operating MCLINC are divided by four. The first 25 per cent of the operating budget is allocated equally among members of the consortium and is called the Membership Share. The second 25 per cent is divided by the number of sites in the consortium (a library counts as a site, and any branches that library has count as additional sites) and is called the Site Share. The final 50 per cent is divided by the total number of workstations in the consortium and is called the Connection Share. Each library pays:
- •
one membership share;
- •
"n" site shares (where "n" is the number of sites for that library;
- •
"n" connection shares (where "n" is the number of workstations for that library).
New England Law Library Consortium
Each member pays a $4,000 flat fee. If the consortium decides to purchase something not included in the existing budget (e.g. a new database), that cost is divided equally among participants.
Novanet Library Consortium
Novanet is a consortium of eight universities in Nova Scotia. Membership fees are based on four factors, weighted equally: student FTE count, university budget, circulation, and the number of records in the bibliographic database.
Old Colony Library Network
The OCLN serves public libraries south of Boston. They use three factors to calculate assessments: 22 per cent is divided equally among all members, 38 per cent is based on the number of ports, and 40 per cent is based on annual circulation.
Pennsylvania Academic Library Connection Initiative
PALCI has a straightforward way of allocating costs among members. Each library, regardless of size, pays a $6,000 fee. This method was used to get the consortium off the ground, and is limited to two years, i.e. each library pays a $6,000 flat fee for the first two years of the consortium's operation. PALCI is now beginning to discuss what cost allocation method should be used beginning with the third year of operation.
State University of New York
SUNY's Office of Library and Information Services supports a Library Automation and Implementation Program (LAIP) that provides library automation services for approximately 40 SUNY campuses, including both state-operated and community colleges. Cost allocation for the LAIP is based on two factors: a flat fee and a variable per-record cost. The flat fee is the same for all institutions and starts at a level of $4,500, increased by $1,250 annual increments up to a maximum of $10,500. For planning purposes, per-record costs were calculated at $0.06 for each record in the bibliographic database.
South Central Research Library Council
This consortium (with headquarters in Ithaca, NY) bases membership fees on the operating budget of the library. They have established three categories, based on size of operating budget. A library's membership fee is dependent on which category it falls into.
South Dakota Library Network
SDLN is a multi-type consortium with 47 member libraries. Their membership fees are calculated by starting with a base fee of either $2,000 or $5,000 (base depends on size of library and services used). The remainder of the membership fees (i.e. the difference between operating budget and total base fees) is prorated using a weighted formula based on the following:
- •
20 per cent of the proration is based on the institution's relative percentage of total bibliographic records.
- •
20 per cent is based on institution's relative percentage of total item records.
- •
10 per cent is based on institution's relative percentage of total patron records.
- •
50 per cent is based on institution's relative percentage of total system transactions.
SWITCH Library Consortium
This Milwaukee-based academic library consortium divides costs according to the following model: 10 per cent of the costs are shared equally by all members, with 90 per cent of the costs subject to a formula as follows:
- •
30 per cent based on collection size
- •
30 per cent based on user community size (including faculty, staff, student and community patron records)
- •
30 per cent based on the number of "access points" (i.e. ports).
Tampa Bay Library Consortium
The Tampa Bay Library Consortium runs an automated library system (SUNLINE) that is shared by 14 TBLC members. Membership costs for SUNLINE range from $7,000 to $30,000, depending on the size of the library.
TexShare
TexShare has two methods of allocating costs. For programs subsidized by centrally provided TexShare funding, TexShare pays 75 per cent of the costs, and the members pay for the remaining 25 per cent. That 25 per cent is allocated among member libraries based on their collection size or student FTE count. A library is placed in one of three tiers based on collection size or student FTE, and then pays the fee associated with that tier. For non-subsidized services, libraries are charged a close approximation of the direct costs of the service to the library.
University of Maryland System
According to Ron Larsen of the University of Maryland System, they initially considered eight to ten different factors "expecting to see a useful pattern". No pattern emerged, so they developed an ad hoc set of percentages to apply to each institution, based loosely on the size of the institution.
Washington Research Library Consortium
When the DC-area WRLC first established its cost-sharing formula, it used the following model: 25 per cent of the budget would be divided equally among the members (originally eight research libraries, now seven), 75 per cent of the budget was prorated according to a formula that considered volumes held, volumes added, serial subscriptions, total library expenditures, and library staff size (FTE). In 1992, when the WRLC was reorganized, the percentages in effect at that time became the official percentages used to allocate costs on an annual basis, with the formula used to calculate the percentages to be revised at least every five years, beginning in 1997.
The 1997 review resulted in a significantly different formula, using some basic guiding principles: the formula must reflect usage of WRLC resources and contribution to resource sharing, must be structured to encourage new membership, and should provide stability and predictability from year to year. Under the new formula, the WRLC budget is divided into two categories:
- 1.
administrative/infrastructure costs (20 per cent of budget); and
- 2.
system operation costs (80 per cent of budget).
The administrative/infrastructure portion of the operating budget is divided evenly among members. The remaining 80 per cent of the operating budget is divided into two categories, and is prorated as follows:
- 1.
60 per cent of the operating budget by a Usage Index, based on student/faculty head count, and library staff FTE.
- 2.
20 per cent of the operating budget by a Contribution Index, based on library materials budget and volume count, inverted so that larger contributors pay a smaller share.
A comparison of the methods used by these 21 consortia reveals some interesting patterns and commonalties:
- 1.
I identified nine factors that these 21 consortia used in allocating costs:
- •
Equal division of costs
- •
User counts
- •
System usage
- •
"Ports" (i.e. some count of the number of terminals/workstations accessing the system)
- •
Collection size
- •
Library operating budget
- •
Library materials budget
- •
"Size" of institution
- •
Number of sites
- •
- 2.
Most consortia use a combination of measures when allocating consortial costs, with only slightly more than one-third of the 21 consortia employing a single measure.
- •
eight consortia employed just one measure
- •
four consortia employed two measures
- •
seven consortia used three measures
- •
two consortia used four different measures
- •
- 3.
The most popular single measure employed by these consortia was "user counts" (e.g. student FTE counts for academic libraries, population served counts for public libraries, etc.). Nine consortia employed "user counts". In one case, a consortium used user counts as its sole method of allocating costs.
- 4.
The next most popular measure was to divide some portion of consortial costs evenly among member institutions. Nine groups employed this method. In two cases, consortia used this measure as its sole factor in dividing costs, meaning each participant paid the exact same share of the cost. In some cases, consortia charged a flat rate fee to cover part of the consortial costs, and used other measures to allocate the remainder.
- 5.
The third most popular factor was "collection size". Seven consortia used this factor. No consortium used collection size as a single factor.
- 6.
The fourth most popular factor was "system usage", i.e. some measure of relative system usage. Six consortia employed this method. No consortium used system usage as a single factor in allocating costs, which is interesting, as a case could be made that relative system usage, where available, might be considered the "purest" factor: the institutions that use the system the most pay a larger share of the costs.
- 7.
The fifth most frequently used factor was the number of "ports", i.e. the number of terminals or workstations connected to the system. Five consortia used this factor. No consortium used it as a single factor. This particular factor will become less and less relevant over time as the concept of "port" loses relevance in the Web-based environment and workstations become less likely to be directly or permanently connected to a system.
- 8.
The remaining four factors were used less frequently.
Conclusions and considerations
If the trends indicated in this study are representative of how consortia divide costs, consortia should consider starting with the following generic model:
- 1.
Begin with a flat base fee that is equal for all member institutions. This fee should be large enough to be seen as a significant contribution, but small enough to be affordable by the smaller or "poorer" institutions.
- 2.
Divide the remainder of the costs using multiple factors, weighted so that they make the most sense in the specific environment. I would recommend considering the following three factors, in descending order of importance:
- •
Relative system usage
- •
User counts
- •
Collection size.
- •
If any lessons can be learned from this study, it might be suggested that most consortia should employ more than one measure when dividing consortial costs. Very homogeneous consortia, with libraries of the same type and size, might be able to use a single factor in allocating costs, but the more diverse a consortium is, in terms of types and sizes of libraries, the more important it is to consider using a mix of factors in cost sharing.
Bernie SloanSenior Library Information Systems Consultant, University of Illinois Office for Planning and Budgeting, Urbana, IL