APPENDIX A
INTERNET COST MODELS FOR PUBLIC LIBRARY NETWORKS
STATE AND REGIONAL PUBLIC LIBRARY NETWORK COST WORKSHEETS
STATE AND REGIONAL PUBLIC LIBRARY NETWORK INTERNET CONFIGURATIONS AND SERVICES
APPENDIX A.1 - NETWORK PARTICIPANT LETTERAPPENDIX A.2 - COST WORKSHEET GLOSSARY
APPENDIX A.3 - COST WORKSHEET (in PDF format)
The initial analysis of data from the 1997 ALA/NCLIS Survey of Public Library Internet Use indicated some difficulties in the data collection specific to survey questions 16 and 17 (see Appendix B):
Public libraries had considerable difficulty in responding to these questions if they were part of a larger regional or statewide public library network through which they received Internet-related services. Many individual public librarie s did not have access to these parent organization budgets and thus, could not answer questions 16 and 17 of the survey. The survey sample did not account for the role of regional and statewide public library networks in supporting local libraries' Intern et access. Since a major intent of the survey is to capture Internet-related costs to public libraries, it became important to better identify, define, and understand these costs related to public library networks.
It was not surprising that public libraries could not, in many instances, provide data relating to Internet costs. This issue constituted a major finding in a study of Internet cost models and public libraries, Internet Costs and Cost Models for Pu blic Library Internet Use (McClure, Bertot, and Beachboard, 1995). That study found that while many public libraries reported they were connected to the Internet, the libraries could not identify and describe specific costs for providing Internet-rel ated services. Other public libraries indicated that connecting to the Internet was a goal, but they were uncertain as to what types of expenditures this would involve.
This inability to predict Internet-related costs was a major stumbling block to many libraries that were not connected as it made any planning efforts difficult if not impossible. But the 1995 study also pointed out that while cost accounting
was a stumbling block, the public libraries saw the need to develop a
methodology that would capture Internet-related expenditures. The study
noted that "cost factors were extremely important to determining the
public libraries' connectivity to the Inter net." (p. 1). At stake here is
the larger issue of the role of public libraries in the National
Information Infrastructure, with which the cost data is integrally
aligned.
Thus, the study team proposed and developed a case study to obtain additional cost data to better understand cost categories and actual costs. Severe time constraints limited the data that were reasonable to collect and the number of public library net works that could be included in the case study. The purpose of the case study was to provide additional descriptive data of cost variables relevant to statewide and regional public library networks to better define these cost variables in a number of diff erent networking environments.
The preliminary response to the 1997 survey reinforced the difficult nature of this task. It is not unusual to experience ambiguity and uncertainty whenever trying to capture initial costs of a new technology (Caudle, et al. 1989; Fletcher et al, 1992) . This does not however, make that process futile. Rather, it argues for the immediacy of such a task in establishing standards at an early stage of diffusion to better track costs incurred and to plan for future development. While the 1995 study made not e of the significant difficulties public libraries had in attempting to define cost categories and provide cost data for Internet-related services, it did identify a preliminary model for cost categories. This preliminary cost model added to the knowledge base of cost information for Internet connectivity for public libraries. The goal of this current study is to build upon that knowledge base, using the preliminary cost model as a starting point to better capture accurate and valid cost data related to I nternet connectivity.
The effort to explore and identify library Internet costs required the use of multiple methods. This section details the methodologies used by the study team to further determine library-related Internet costs with a selected set of libraries
and library systems representative of existing Internet connectivity
models --statewide, regional, and individual library systems.
Research Questions
The case study portion of the research is a supplement to the 1997 ALA/NCLIS Survey of Public Library Internet Use. The study team developed the following set of research questions to better understand the degree to which public libr
aries received subsidized allocations from statewide and regional pubic library networks to provide Internet services (survey questions, 16 and 17):
In addition to these specific questions, the study team used this opportunity to gather data on statewide and regional public library network configurations and Internet-related services. This secondary data analysis serves to provide a br oader picture of the relationship of these networks to public libraries in provision of Internet access.
Preliminary Cost
Model
The authors developed a preliminary cost work-sheet which was
taken directly from the worksheet presented in the Internet Costs and
Cost Models for Public Libraries (McClure et al., 1995). The study
team assessed this cost worksheet for its utility and currency to the
Internet environment of 1997. A number of the line items were developed
based on changes in network technologies since the time of the original
worksheet. Data obtained from the 1996 Public Libraries and the
Interne t study (Bertot, McClure, and Zweizig, 1996) and other
projects in which the authors were involved ( Bertot and McClure, 1996;
McClure and Bertot, 1997) further informed the cost model revisions.
The study team added a new category--telecommunications lines to the Internet--to the pre-test version of the cost worksheet. Line items added to this category included dial-up and leased lines for added detail. The original cost worksheet
section of "soft-ware" was expanded to include Web page expenditures and
software for people with disabilities. The original "content/resources
development" category came under revision to reflect the electronic nature
of content and resources that the s tudy team wanted to capture.
This revised version of the original cost worksheet became
the instrument for the pre-test portion of the case study. The pre-test
asked case study participants to evaluate the usability of the revised
cost model and to make changes to catego ries and line items as they saw
fit. Three iterations of this process were con-ducted to produce a "final"
version of the cost model to use for the case study.
Data Collection
The study team collected data from July 10, 1997 through August 27, 1997. The team developed a case study methodology for the collection of data and employed a variety of data collection techniques to answer the research questions. Initially,
the study team reviewed preliminary survey data reported in the 1997
ALA/NCLIS Survey of Public Library Internet Use. In specific, an
analysis of survey questions 16 and 17 framed the collection of data from
the case participants. This analysis included discussion of key variables
with case study participants and the Advisory Board to the study team.
In addition, the study team reviewed the data from the
above-mentioned cost model study sponsored by the U.S. National Commission
on Libraries and Information Science (NCLIS) (McClure, Bertot and
Beachboard, 1995). The cost worksheet develope d for that research became
the basis for cost worksheet development in this case study.
The study team modified the cost worksheet to reflect
subsequent changes in information technology to better capture a number of
different expense items. Once the initial cost worksheet was revised and
finalized, a timeline was developed for collecting the data. As this
portion of the larger study emerged from preliminary data analysis of the
survey, time was a limiting factor in the data collection effort. The
survey portion of the study was originally scheduled to be completed by
May 23, 19 97. The study team extended this deadline to allow for greater
participation from the public library sample. There remained, however, an
overall timeline for completion of the study that did not allow for
significant deviation. Thus, the case study needed
to be completed within a six week timeframe.
The study team contacted potential case study participants by
e-mail and fax to introduce this portion of the study and solicit their
interest and willingness to participate. Four of the original nine sites
contacted were unable to participat e due to the severe time
constraints.
The remaining five participant sites were first asked to pre-test the revised cost worksheet. Their comments were used in further revisions to the worksheet and in the development of specific interview questions which would elicit responses to the rese arch questions posed above. The study team finalized the questions and faxed them to all participants on August 5, 1997. The fax contained a cover letter (see Appendix A.1) with specific requests for information and the fourth iteration of the cost worksh eet for their analysis.
At this stage of the research, the study team determined that it would not be possible for any one public library network to use some standard cost worksheet to capture their Internet-related costs. The discussions held with the participants
during the pre-testing of the cost worksheet indicated that rather than having one standard cost worksheet, multiple versions of such a worksheet would be necessary, tailored to the budget processes used in the various jurisdictions. Instead of developing
multiple worksheets to try to accommodate local differences, and then having the case study participants fill in the cost data for these, the study team asked the participants to respond to four questions pertaining to the current revised cost worksheet.
The participants received a deadline of August 15, 1997 in which to
respond to the research questions either by phone, fax or e-mail. Along
with this set of interview questions was a list of documents deemed
relevant to collect for the research. The case
study participants were requested to submit any documents relating to their network configurations, information architectures, network policies, and listing of network services.
The study team, in the development of this portion of the
larger study, used accepted techniques and methods for the development of
the research design and data collection and analysis activities (Creswell,
1994; Denzin and Lincoln, 1994; Marshall and Rossman, 1995). The study
team used specific strategies to insure the reliability and validity of
the data collected. It is important to note that the small size of this
case study does not lend itself to generalizing the results reported her
e. Rather, the data should be viewed as a preliminary attempt to describe
the range of issues and factors confronting statewide and regional public
library networks when public libraries attempt to accurately cost their
Internet services.
Case Study Participants
The study team first reviewed the preliminary returns of the 1997 ALA/NCLIS Survey of Public Library Internet Use to determine if certain statewide or regional public library networks were identified. The study team then conducted
a search on the Internet to identify statewide and regional public
library networks using the following URLs:
<http://sunsite.berkeley.edu/libweb/usa-pub.html>;
<"http://rio.atlantic.net/~bdarl/library.html.plibrary>
;<http://sjcpl.lib.in.us/homepage/PublicLibraries/PublicLibraryServers.html>.
The study team was able to obtain descriptions of these networks from a search of their Internet addresses. This gave the study team a preliminary list of potential case study participants.
The study team established the following criteria to identify appropriate statewide and regional public library networks for participation in the case study:
In all, the study team collected data from three regional public library networks and two statewide public library networks.
The study team used the following techniques to collect
data:
This combination of data collection techniques enabled the study team to obtain as rich a picture as possible from a range of networks within a limited timeframe. The various data collection techniques were useful in informing the revision s to the original cost worksheet sent to the participants. In all, the study team revised the cost worksheet (Appendix A.2) three times in response to feedback from the case study participants. In addition, the study team developed a glossary (Appendix A. 3) based on requests for information from the participants to clarify the line items on the cost worksheet.
Data Analysis
At various points in the case study, the study team reviewed
the participants' comments for the purpose of revising the cost worksheet
and further articulating the data needed to answer the research questions.
At the end of the second round o f comments from the participants on the
feasibility of completing the cost worksheet for their network and on the
usefulness of the cost worksheet to capturing their Internet expenditures,
the study team revised its goal of collecting statewide and region al
public library network cost data by having each participant complete a
cost worksheet. The feedback from the case study participants indicated
that due to the complex and diverse nature of their networks, budgeting
methods particular to each site, and other differences in service range,
no one cost worksheet would be adequate to meet the needs of all statewide
and regional public library networks.
Based on this feedback, the study team prepared a "final"
version of the cost worksheet for the parti-cipants. The participants were
requested to respond to a subset of questions directed towards
understanding the overall issues of costing ou t their network services in
connection with the use of the final version of the cost worksheet.
The participants submitted their responses via e-mail, fax
and telephone interviews which were, in turn, analyzed for content issues.
Documentation about costs and configurations accompanied this data and was
content analyzed to supplement th e interview data. It should be noted
that not all respondents provided data for every data collection request.
Thus the analysis of the data reflects the responses received that will
not in all cases add up to five.
STATE
AND REGIONAL
PUBLIC LIBRARY NETWORK COST WORKSHEETS
The data presented here come from the
pre-test of the cost worksheet, phone interviews about cost worksheet
issues, and a set of questions sent to all participants to ascertain the
following:
Below are the issues and results study participants provided concerning the worksheet.
The Cost Worksheet
Download a copy of the cost worksheet, in
PDF
format
Case study participants received initial cost worksheets to pretest for usefulness and clarity. Their comments, reported here, were incorporated in to changes made to the worksheet. This final version of the worksheet was then re-sent to the participants to assess its applicability and utility.
An overall theme from the respondents was that the
expenditures could not be broken down as neatly as the cost worksheet
assumes. The major problems with the pre-test version of the cost
worksheet included the inability to:
The study team took these comments into account and
revised the pre-test version of the cost worksheet to a version that more
accurately represented the budget concerns of the participants. The
"staff" category, section 8 on the cost works heet, was significantly
revised to enable respondents to indicate multiple and unique staff lines
including a section to "check-off" the staff participation in training,
help functions, planning, and content development functions.
The "no-cost item" category, section 9 of the worksheet was
revised and simplified to enable respondents to merely indicate if they
received any of the named resources rather than requiring them to attempt
to cost out the value of these items . Another direction taken by the
study team was to develop a glossary of operational definitions for the
line items as suggested by the respondents. This glossary accompanied the
revised cost worksheet in the final round of data collection.
Most Difficult Issues
The case study participants defined the three most difficult
issues they faced in completing the latest version of the cost worksheet
given their divergent network configurations, budget methods, and
services. A major theme noted by the parti cipants was the difficulty in
defining what, exactly, was a network. There was no common understanding
of what constitutes a network and this made completion of any
type of cost worksheet problematic. A related issue here is the rapidly
changing network technology and policy environment. This lack of stability
makes defining and tracking costs again, problematic. Participants
repeatedly noted that the transitional nature of the network environment
was a barrier to accounting.
The difference between the cost worksheet and the accounting
systems used by different jurisdictions was another noted difficulty. To
make the cost worksheet viable, one respondent commented that she would
have to set up a separate spreadshee t to capture the expenditures listed
on the cost worksheet, and another spreadsheet to capture the expenditures
as required by their budget and accounting system policies.
Another noted difficulty was the problem of tracking shared resources or "free" resources in terms of both costs and quantity. The respondents stated that a number of the expenditures on the cost worksheet are not used solely for the network, but are a lso included in other library services, e.g., consolidated telephone bills are not amenable to calculating the long distance charges directly attributable to Internet activities. Another example of a shared resource was software where the license was for multiple simultaneous users, not all of whom used it for Internet activities. In some cases, resources, e.g., equipment and staff, are shared with other agencies and would be impossible to determine how much should be charged to one agency versus another. A good portion of such information was "not available, not predictable, and not under my control."
One respondent noted that the accounting practice used by
the library network is to aggregate some of these items and that other
cost items are shared in such a manner that makes their documentation
difficult. Among the most common expense s that are paid out of other
funds or provided by other agencies are software licenses, telephone
lines, Internet access, workstations and servers. Some networks receive
these items as "in kind" contributions, others are bundled with other
services, and s till others are purchased and/or maintained by parent
organizations.
Respondents noted further that a number of the line items on the cost
worksheet were irrelevant as the items represented one-time costs for the
network that had already been expended. In one instance, the library
network purchases workstations on a sta ggered basis, with 25 percent of
all workstations replaced each year. Participants commented that tracking
some of these expenditures over time would not be a useful process.
An important consideration to note is that the cost
worksheets are political by their very nature. Introducing and
acknowledging this element of "politicalness" makes problematic the
completion of any worksheet. Actual expenditures for tec hnologies are
often (purposefully) embedded in other cost categories and reporting these
costs might better stay hidden. Budgets are political tools and cost
categories become the vehicle for the playing out of political agendas in
an agency. The expenditures noted in any budget say more than a dollar
value. They reflect choices, philosophies, and policy, along with
alliances and compromises made to further the goals of an agency. These
implicit variables are subtle, and forcing them to become explicit is
not only futile but foolhardy in the eyes of agency executives.
A final concern noted relates to the fact that the network
infrastructures of the participants cover a range of installations and
agencies, e.g., member libraries, state library agencies, county
governments, consortia, and the state or region al network organization.
This multiple cost structure makes it difficult to provide cost data. To
do so would require that the member libraries each fill out a cost
worksheet and submit it to the parent network organization. Respondents
noted the difficul ty they would have in estimating the expenditures of
these network items for their member libraries.
An additional complication here is that the member libraries do not solely use the infrastructure for Internet access. Other uses are made of the infrastructure, including automated circulation, inter-library loan, and public access to databa
ses. Attempting to factor out these other services would be, if not impossible then, very difficult and may not generate accurate cost estimates. A final complication occurs where other agencies "bill" the state or regional network for services which the
networks cannot directly contract, such as voice or data telecommunications. In these instances, the billing does not correlate with any of the cost categories on the cost worksheet.
Time to Complete the Worksheet
There was a wide range in the estimates given for the time it
would take to complete the cost worksheet as received. One respondent
stated that she "hadn't a clue" and even if she did, for reasons mentioned
above in the "difficult issues" sec tion, the information would not be
meaningful to their network. Other responses ranged from six to eight
hours; eight to ten hours; and twenty-four to thirty-two hours of time. It
is important to note here the comment from some of the respondents that
reg ardless of the time allotted to completing this cost worksheet, they
were unsure from the start that they could obtain the required data.
The "twenty-four to thirty-two hours" response was from
one of the statewide public library networks. The respondent further noted
that even if they standardized the administration of this cost worksheet
for their statewide public library network, that lengthy amount of time
would not likely be reduced. The respondent noted further that even given
the thirty-two hour time frame, the cost worksheet would still be
incomplete as some of the data requested on it could not be provided due
to th e extensive amount of additional time it would require to extract
that information.
Another issue raised here is that some of the participants
remained unconvinced of the value of determining the costs given the
effort required to capture them. It is important to raise this concern of
effort versus value. One respondent p ointed to the extreme level of
detail requested in the cost worksheet, and wondered how important it was
to have separate cost data for line items such as routers, modems, number
of PCs, or operating systems.
Cost Worksheet Modifications
There were a range of responses to the request to provide
suggestions to make the cost worksheet more useful to each participant's
network. In the case of the smaller regional public library networks, the
responses concerning modifications we re not drastic. One respondent noted
that few modifications were needed as they would be able to fill in the
worksheet fairly easily as it stood. Another regional public library
network noted that most of the modifications they would require were in
Secti on 2, "Telecommunications Lines Out to Internet". Here they would
need to expand the line items to better reflect their specific
telecommunications infrastructure which enables all their member libraries
(43) to connect to the Internet.
One respondent made the comment that it was too early in the
Internet era to talk about modifying the cost worksheet as the situation
right now is too dynamic for any one stable document to capture. The
number and type of Internet-related ser vices is expected to change.
Pending and future policy decisions will change how cost information is
captured, and what cost information will be tracked. The resolution of the
FCC's telecommunications discounts will also affect what and how they
track cos t data (see the FCC's website for additional background
<http://www.fcc.gov>).
A respondent from one of the statewide public library
networks provided a number of cogent and detailed comments to the request
for suggestions to modify the cost worksheet. First, she noted that the
revised cost worksheet had considerably mo re clarity than did previous
versions. The "personnel" section was mentioned in that it now asked for a
doable and sufficient level of detail for which to provide costs. A
suggestion was made to reinstate the "travel" line back into the worksheet
with a r ecommendation to place it under Section 7, "Grant/Program
Development". The study team deleted the travel line along with the other
Personnel lines (staff training, user training, documentation development,
on-going user support) after comments from the p re-test indicated that
this level of specificity was difficult to provide given the sharing of
personnel resources and the volatility of the current personnel
environment.
One respondent went on to note that to make more detailed modifications to the current cost worksheet might not be appropriate for other public library networks as the accounting system they used might not be transferable to other jurisdictions. One su ggestion to ameliorate this concern would be to collapse some of the line items to reflect a broader level of detail, rather than the fine level of granularity the cost worksheet currently attempts to capture. Suggestions made included combining LAN hubs, routers, modems and transceivers in Section 3 of the worksheet, and combining a number of the software costs as the level of detail in section four seemed like "overkill." A suggestion made for Section 6, "Electronic Content", concerned the issue of Web- based content versus all electronic resources a library might offer. The respondent noted that further clarity in this section would be useful to ascertain if the line items reflected only Web-based services or all electronic services.
Annual Cost Data
As anticipated, it was very difficult for the case study
participants to provide the annual cost data since the inception of their
networks. A number of the respondents were unable to provide any data for
a variety of reasons. A common theme heard here was the problem presented
with overlapping staff functions. Library staff in many instances were not
dedicated full-time to Internet-related work making it very problematic to
calculate percentage of time and cost to their various functions. An other
difficulty the case study participants had was in the inability to
calculate the dollar amount that their "free" access or equipment
represented. Some Internet connections were provided through other venues
which absorbed the cost and thus, the publ ic library networks would not
have an actual dollar amount to report.
An additional concern raised is the recency of costing out network expenditures as separate from the more traditional information services function. One statewide public library network participant noted that prior to Fiscal Year 1996 their Network Ser vice pro-gram was not a separate organizational budget code for them. Supplies, printing, training costs, and general personnel expenditures had been mixed in with the Information Services budget. This type of budgeting is not unusual for the other case s tudy participants, as noted in their recurrent comments on not being able to cost out from their general budgets these Internet-related expenses.
Where dollar amounts were provided by participants, there was
an expected range of annual costs. The size of the networks, the range of
services, and the limitations expressed about capturing many of these
expenses account for the differing a ggregate costs. One regional public
library network provided an annual network cost estimate of $150,000 -
$200,000. Another regional respondent gave $862,000 as a 1996 actual
amount. One of the statewide respondents provided a figure of $1,167,008
for ex penditures to date since the inception of the network in 1992. The
1996 expenses provided by this network totaled $372,234.
In two cases, the participants did provide a level of detail
of expenditures over the network life. The participants did note that what
they provided was problematic in that their budget methods tracked
different and in some cases cross-funct ional budget line items so what
they submitted reflected their best effort to break out the cost. Their
worksheets are presented in Figures A.1 and A.2.
There is an inherent difficulty in understanding of these cost estimates. An estimate provided by a regional network is low due to the large amount of in-kind, shared, and "free" services they receive from other organizations. In other cases,
start-up expenditures or one-time expenditures can artificially inflate a
cost estimate. These issues, coupled with the uncertainty the respondents
had in interpreting the cost worksheet line items, make any estimates
tenuous at best.
Summary of Cost Model Issues
There are some major and recurring themes relating to the
usefulness of the cost worksheet at this time in the development of public
library networks. First, the fact that public libraries are in a
transition phase with network development, u se, technology, and policy
makes accounting for a routinized subset of expenses all but impossible.
Until such time as this environment stabilizes, the development of any one
standard cost worksheet is not useful. But the development of individual
cost wo rksheets reflective of a jurisdictions budget and accounting
procedures is useful and necessary to track current Internet-related
expenditures and to plan for future Internet development by public
libraries.
Another concern for cost modeling is the vastly differing
nature of how the statewide and regional pubic library networks budget.
Line items, accounting codes, and tracking procedures differ significantly
from jurisdiction to jurisdiction. Th is makes standardization a
challenge.
A third issue is the inability to cost out those expenses
that are shared over multiple library programs or bundled with related
services. Because of the idiosyncratic nature of how each network
allocates and tracks resources, it is a challen ge to determine what
portion of a resource is actually an Internet cost versus a circulation
cost or an administrative cost, for example. The fact that these resources
are often shared over multiple programs, and not dedicated solely to
Internet use, make s accounting for the resources an inaccurate science.
But without any reasonable estimates of these costs, planning,
rationalizing, and accounting for Internet-related services will suffer
from a lack of justification.
STATE
AND REGIONAL
PUBLIC LIBRARY NETWORK INTERNET CONFIGURATIONS AND SERVICES
As an addendum to the cost model portion of the case study,
the study team requested additional documentation from the participants
that described the configuration of their network and the services their
network provided. Any existing networ k architectures or blueprints, and
any description of services provided by their networks were also requested
by the study team. As with the cost worksheet portion of the case study,
the participants' ability to respond to this secondary request ranged fr
om the provision of actual blueprints to comments reflecting the
respondents' awareness of the importance of such a document and their
lament that, at this time, they did not have such.
Regional Public Library Networks
The scope of each regional network varied:
The frame relay network utilized by the three regional networks can support speeds from 56 kbps to 1,544 mbps. Both networks are ethernet with Cisco 7000 or x.25 routers.
One network has been offering dial-up access to the Internet through terminals in the public libraries or through 16 dial-in lines since 1993. A second regional network initiated Internet access in March, 1997, via a 10 workstation computer l
ab. Netscape and Internet Explorer are the main Internet browsers used. Although Lynx is also offered for a text-based Internet component at one of the regional networks.
The regional networks offer a range of services at their web
sites. One of the regional networks offers local community and government
information and use of periodical and newspaper databases. Another makes
available a variety of software to ols, e.g., Word, Excel, and Power
Point, Encarta 97 Encyclopedia and World Atlas, a database of full text
articles and images from periodicals, newspapers, directories, handbooks,
and the like, as well as many multimedia and interactive databases.
Interne t training - beginning and advanced - is an additional service for
the library patrons one of the regional networks. All three regional
networks have their library catalogs on the web.
An interesting feature of the Internet-related services is the partnering which describes some of these ventures. Through a "Libraries Online" grant, one of the regional networks created the "Power Up Place." This computer lab is a cooperativ
e effort of the Microsoft Corporation, the American Library Association and the Technology Resources Institute. Another partnership noted is between one of the regional networks, a regional Internet service provider, and a private training company to offe
r "affordable, user-friendly Internet classes" for library patrons.
A third partnering arrangement occurred between one of the
regional library networks and Bellcore for the purpose of investigating
public access to the Internet. A pilot project was launched to determine
public and institutional use of the In ternet with a study group of 200
library staff and 174 library patrons. The project ended in August, 1995
with the private sector involvement terminating. The library staff and
patrons chose to continue the Internet access after seeing the value of it
as an information resource.
Only one
regional network provided data on governance of their Internet-related
services. They are governed by a 14 member, elected, Planning Council. The
members are composed of 12 library directors plus two county employees,
the County Libr arian and the Chair of the Department of Community and
Special Services. The Planning Council, in turn, reports to the County
Freeholders on network services and issues.
Statewide Public Library Networks
It is difficult to summarize the
scope and services offered by the two statewide public library networks
here and capture the full flavor of what they do and how they do it. Much
of the information presented here came from their web sites, su pplemented
by press releases, articles from periodicals, and information obtained
from email communications with the library directors. Each statewide
public library network will be reported on individually.
The first statewide network officially began
with a state legislature appropriation in 1992. The appropriation provided
$50,000 for one-time networking activities by public libraries. Eight
pilot sites were chosen for the initial funding. Sin ce that time, with
additional appropriations from the state legislature, 44 libraries serving
approximately 90 percent of the state population have begun network
activities.
Other activities initiated by
the state library include:
This statewide public library network was a 1995 National Information Infrastructure Award winner, and is a nominee for a Service Application award from the National Association of State Information Resources Executives.
The second statewide public library network was proposed in
1994 to the state's Information resources management Commission. The
stated goal was to:
provide all of the state's citizens equal and unimpeded access to extensive libraries, leading research facilities, government agencies, powerful computing centers, and a wide-range of government and private services. Librarians are already playing a k ey role in enabling citizens to access information through electronic networks. As the sources proliferate and the technology becomes increasingly complex, librarians will continue to expand their role in assisting users to cope with a sophisticated techn ical infrastructure, discover vital information resources, and take advantage of the many applications available for locating and retrieving information. The establishment of a Network Information Center by the State Library embodies this expanded role fo r librarians.
A web site was developed with information resources relating
to genealogy, state government documents, state newspapers, statistics,
and public libraries on the web. Email is provided for public library
staff statewide as are some listservs. Training in Internet-related
skills and consulting for technical assistance are also provided to the
public libraries.
It is clear in examining these differing services that public
library networks are still testing the waters on the Internet. The
networks made clear their desire to assist the public in accessing the
best information resources through a varie ty of network activities. These
activities are still being thought of and developed and will continue to
emerge as new creative and productive ideas evolve.
It is clear from the results of the
case study that capturing costs and developing services on the Internet is
akin to the PC revolution in the early 1980s. Here was a fantastic new
technology that brought the power of information to the desk top. Yet, the
full capabilities of the technology were unknown, and the effective
capture of the cost and cost savings was not obvious. The emerging public
library networks are in somewhat the same state in regards to Internet
costs and services. What is missing from this scenario is the
fragmentation of data and organizational functions that were an unintended
by-product of PC use. The public library networks, rather than enabling a
pell-mell development of Internet related services, are facilitating a u
nified and cogent approach to developing an Internet agenda for their
regions and states.
A major conclusion from this case study is the finding that making any kind of cost comparisons across public libraries at this point in time is problematic. There are no obvious consistencies in how public libraries track budget expenses, no
r are there any standard budget codes being used across libraries. This is partly a function of the nascency of the Internet, and partly a function of the idiosyncratic budget processes used across differing jurisdictions.
The case study findings demonstrate the difficulties related to identifying, describing, and defining cost data and categories related to public library use of the Internet in a networked environment. These difficulties, do not, however, amel
iorate the importance and usefulness of such data for local planning and for national policy making related to the role of public libraries in the NII. Indeed, an important contribution of this study is increasing the awareness of such
difficulties and issues associated with identifying, describing, and defining Internet-related costs for public libraries. Public libraries need to develop their own mechanisms to capture their Internet-related expenditures. As the services and contingen
cies associated with the Internet become more evolved, it becomes expeditious to cost and account for these. The obvious role of the budget in a planning process also necessitates that public libraries become more capable in capturing these costs.
The cost worksheets developed from the case study can be used by public libraries and their networks as a first step in identifying, describing, and defining costs and cost categories at their individual libraries. The wor
ksheets can be modified by individual libraries, statewide and regional networks to better meet their needs and situational conditions. While the study team concludes that collecting such data is still difficult, the effort is no less important and librar
ies should continue to experiment with ways to collect and define such data. Indeed, the process of individualizing the worksheet to a particular library or library network will provide significant insights for staff at that library about Internet-related
costs.
For individual libraries attempting to modify the worksheet developed here, the study team would suggest that it is better to have estimates of networking costs based on the best information available than to have no cost data available. Thus
, continued work in this area from both local libraries and researchers is needed to develop more accurate and explanatory models of Internet costs for public libraries. Such research must be continued and expanded -- especially in the context of better u
nderstanding the linkages between Internet costs and Internet services provision. While standardized worksheets appropriate for a range of public libraries and networks may not be possible at this time, it is possible for individual libraries to identify,
describe, and define costs that make sense for them, in their particular situation.
Moreover, describing and defining Internet related costs will become increasingly important as libraries continue to increase the range of services provided via the network. Public libraries need to become more aware of how to manage the inte
grated nature of the information and communications technologies when assessing their infrastructure. As many respondents' noted, costs are often embedded in more than one function, which currently makes accounting difficult. A more "information resources
management" approach might help to better allocate the various costs to Internet-related services. This also holds for accounting for Internet-related resources that are "in-kind' or "free." Somewhere there is a cost for these and a better attention to d
efining this cost will give a more valid picture of total Internet-related expenses. All these costs need to be integrated into the overall costing and budgeting process of the library through its Management Information System (MIS) or its Decision Suppor
t System (DSS). Regardless of the approach taken, comprehensive management of library costs is essential for successful planning and evaluation of library services.
To a large degree, public librarians, researchers, and policy makers do not understand the linkages between Internet costs and the quality and type of Internet services that can be provided given a specific level of costs. Clearly, low bandwi
dth will limit the types of networked services that can be provided by a library. But the degree to which other cost factors directly affect quality and type of service provision is unclear. Costing Internet-based services and comparing such services cost
s to the more traditional services costs will continue to be an important area for future research.
The study also indicates the importance of resource sharing as a means for reducing overall networking costs on an individual library basis. In fact, the success of such efforts is indicative in the inability of libraries and the networks to
be able to parse out specific Internet-related costs. But evidence from this study suggests that there are important economies of scale for those libraries participating in statewide or regional/local networks in lowering overall Internet costs. A better
understanding of these economies of scale is essential for "next steps" and innovative models for statewide and regional resource sharing.
Internet-related costs for public libraries are likely to increase as a percentage of the overall library expenditures in the near-term. Thus, issues of identifying, describing, and defining those costs are not likely to go away any time soon
. The worksheets and suggestions offered in this portion of the 1997 Public Libraries and the Internet study can provide libraries and policy makers with a good first step in how to best go about estimating those costs. Such estimates will be ess
ential in managing networked library services and insuring that the services provided in this networked environment best meet the needs of the users of these services.
There is the necessity to capture Internet-related costs as they pertain to a performance-based environment. Until the costs can be adequately captured, evaluations will not be able to be tied to program expenditures. The will of the electora
te is reflected today in the mandate for performance-based results from government agencies. This initiative transcends sector in its applicability. Performance-based results are directly tied to budget appropriations, thus, the need to focus on Internet-
related costs and their requisite services.
The scope and levels of Internet-based services are evolving at different paces across the public library networks. The advanced and exemplary state of some of these networks provides public libraries with strong role models, thus discouragin
g the tendency to reinvent the wheel yet again with the Internet. Experiences with costing, partnering, funding, training, development, and services should be shared across public libraries to foster best practices and organizational learning. The wide va
riation in network membership, range, and scope, rather than being a disadvantage in providing one standard framework, makes possible the development of Internet-related services applicable to a diverse public library network environment.
Finally, it is important to note that the issues pertaining to costing Internet-related services are not for the sole purpose of efficiency. It has recently been suggested that rather than being cost reducers, the information and communicatio
ns technologies actually involve rather large capital outlays and have non-trivial maintenance costs. Relying on information technology to trim budgets may not make the best case for such expenditures nor is it realistic. Start-up costs for the Internet c
an be large, but should not be daunting. The effectiveness measures that can be associated with theses costs make a better case for public library involvement with the Internet. This new channel can enable public libraries to extend their reach to potenti
al patrons, and increase their range of services. The information they can access, provide, filter, and otherwise add value to can only be enhanced by creative use of the Internet. This creative use will in turn help to define the evolving role of the pub
lic library in an information age.
Berkeley Digital Library SunSITE; Libraries on the Web. <http://sunsite.berkeley.edu/libweb/usa-pub.html>.
Bertot, J. C., and McClure, C. R. (1996b). Sailor network assessment final report: Findings and future sailor network development. Baltimore: Md.: Division of Library Development and Services. Available: <http://research .umbc.edu/~bertot/files.html>.
Caudle S. L., Bretschneider, S. I., Marchand, D. A., Fletcher, P.T., and Thurmaier, K. (1989). Managing information resources: New directions in state government. Syracuse, N.Y.: Syracuse University. 307 pp.
Creswell, J. W. (1994). Research design: Qualitative and quantitative approaches. Thousand Oaks, Ca.: Sage Publications.
Denzin, N. K., and Lincoln, Y. S., eds. (1994). Handbook of qualitative research. Thousand Oaks, Ca.: Sage Publications.
Fletcher, P. T., Bretschneider, S.I., Marchand, D. A., Rosembaum, H., and Bertot, J. C. (1992). Managing information technology: Transforming county governments in the 1990s. Syracuse, N.Y.: Syracuse University. 213 pp.
Linkway2; Libraries on the Web. <http://rio.atlantic.net/~bdarl/library.html.plibrary>
Marshall, C., and Rossman, G. B. (1995). Designing qualitative research, 2nd ed. Thousand Oaks, Ca.: Sage Publications.
McClure, C. R., and Bertot, J. C. (1997a). Evaluation of the Online at PA Libraries Project: Public access to the Internet through public libraries. Harrisburg, Pa.: Pennsylvania Department of Education, Of fice of Commonwealth Libraries.
St. Joseph County Public Library: List of Public Libraries with Internet Services.<http://sjcpl.lib.in.us/ omepage/PublicLibraries/PublicLibraryServers.html>.