|
DETAILS FOR COSTING MODELS FOR CONNECTING SCHOOLS TO THE NII
As discussed in the main body of the report, we constructed several models assuming different levels of infrastructure and timing of deployment to highlight the major cost drivers of technology deployment and the economic breakpoints among deployment options. This appendix is for the reader interested in further detail about cost models.
Costing methodology
For each model, we analyzed the costs associated with six elements of infrastructure: the connection to the school, the connection within the school, hardware, content, professional development, and systems operation. Each of these elements was further broken down into sub-elements. (See Exhibit 15: "Six Elements of Infrastructure.")
We took a three-step approach to estimating the costs for each model. First, we estimated the costs of each of the six infrastructure elements (and sub-elements) for an average school(55) as required by each model. For each element, we estimated the costs of initial deployment as well as ongoing operations and maintenance. Initial deployment costs include the purchase and installation of equipment and first-year operating expenses. Ongoing operations and maintenance costs include usage charges, equipment and content upgrades, and professional development and support. For many elements, we assumed that prices would decline over time. We also made adjustments-based on location and age-to account for major variations in costs from school to school (e.g., the greater cost of deploying computers and local area networks in older schools requiring retrofitting and asbestos removal). Second, we estimated the amount and quality of existing infrastructure for each cost element to determine the true incremental costs of deployment. Third, we scaled the costs up to a national level by multiplying the incremental costs per school by the total number of schools, accounting for the growing student population.(56) For each model, we assumed either a 5 or 10 year deployment period (as noted in Exhibit 4) with the purchase and installation of the equipment evenly spread over that period. All costs are in nominal dollars and assume a 3% inflation rate.
Our analysis focused primarily on computer-based infrastructure using networked computers as access devices, though costs were also calculated for dedicated video and for telephones and voice mail. As many industry participants have observed, the distinction among computer, video, and voice platforms will blur as broadband connections become more widely available and as computer technology makes its way into televisions and telephones. Someday, interactive television may rival networked computers as a workable base for connecting schools to the NII. We have focused on computer-based technology because it is widely available today and, therefore, provides a sound basis for cost estimates.
Schools may find they have many connection options, depending on where they are located. These options will include both the medium (for example, wireline options include telephone lines and cable; wireless options include satellite, microwave, and cellular) and the type of service (including bandwidth, features and price) offered. For most schools, we assumed telephone company connections because they are the most widely available two-way connections and, therefore, best lend themselves to pricing estimates. However, because high-bandwidth telephone company connections are not available in all rural areas (or are very expensive), we based some of our models on wireless radio for a portion of schools in rural areas.(57) While satellite, cable, and other wireless connections offer viable and potentially cost-effective alternatives, today only telephone company connections offer full, two-way interactivity to a significant portion of the country.
For purposes of cost analysis, the telephone company connections considered were POTS and T-1 lines. These two offerings represent a limited set of the available services. Individual schools and districts will want to investigate other wide- and broadband services which may be available from the telephone company-including ISDN, frame relay, and LAN interconnection-as well as non-telephone company options. As discussed in an earlier section of this report, alternate services such as ISDN may prove to be more cost-effective.(58) The answer for a given school will depend on its needs, the available options, and the price of those options, all of which vary widely from area to area.
Computer-based infrastructure options
We modeled the technology infrastructure and costs associated with full connectivity in every classroom of every public K-12 school-the Classroom model. We also analyzed three less ambitious models that could be considered as alternative deployment options or as interim steps on the path to classroom connectivity: a Lab model, a Lab Plus model, and a Partial Classroom model. In addition, we considered a Desktop model (one computer per student) but did not focus our attention there, given its relatively high deployment costs.
These computer-based models and their costs are described in several exhibits throughout this report. The key features of each model are explained in Exhibit 3, "Model Features," and the national level costs displayed in Exhibit 4, "Estimated Cost of Deploying and Operating Infrastructure." Exhibit 16, "Model Costs at National Level," shows the breakdown in national costs by element and model. Finally, Exhibit 17, "Different Representations of Model Costs," displays the costs in three ways: national costs, costs per average school, and costs per enrolled student. The costs of the computer-based models are not incremental to one another; this means, for example, that the Classroom model does not include the Lab model.
a. Connection to School
External connection costs include installation, access and usage charges for both the school and the district. We assumed mostly wireline connections (primarily POTS lines for the Lab and Lab Plus models and T-1 lines for the Partial Classroom and Classroom models), although costs for some of the rural schools (27%) were estimated with wireless radio. For example, 50% of the rural schools in the Classroom model were assumed to use POTS lines with wireless radio rather than a T-1 line. We used average current Regional Bell Operating Company (RBOC) tariffs as the basis for cost estimates. Tariffs were assumed to decrease by 3% per year through the deployment period.
As discussed in the body of the report (see Meeting the Funding Challenge), current infrastructure for the connection to the school is quite limited; less than 5% have ISDN or T-1 connections and less than 12% of classrooms have telephones.
b. Connection Within School
Internal connection costs include the materials and labor for installing Ethernet LANs (e.g., cabling and network interface cards) as well as file servers, hubs, and routers. File servers are also included for the district.
Our estimates of the LAN costs varied by the age of the school. The NCES estimates that 65% of schools are more than 35 years old and have not undergone a major retrofit. We assumed that physically wiring these schools would require asbestos removal and other retrofitting (for the Partial Classroom and Classroom models). Given the high cost of such remediation, we assumed that wireless LANs were employed where possible, which we estimated to be half of the schools.(59) The cost of installation for wireless LANs is expected to decrease over the next few years to about $200 per node, directly comparable to wireline solutions. For the other half of older buildings, we assumed $63,500 per school for asbestos removal and additional retrofitting. New schools (5%) were assumed to have adequate wiring already built in. Another 30% of schools are between 5 and 35 years old; we assumed these schools neither had wiring nor required asbestos removal.
We assumed a 10 mbps Ethernet LAN that then shifts over time to a 100 mbps LAN at the same cost. The Lab model includes a server at the school ($3,200) and a server at the district ($10,000); the Classroom model includes 3 servers ($3,200 each) at the school and 2 at the district ($10,000 each).
Based on our review of survey data, we estimate that 7% of classrooms were connected to an Ethernet or comparable LAN in 1994-1995.(60)
c. Hardware
These costs include multimedia-capable computers, printers, scanners, furniture stations, and security systems. They also include any facility upgrades or retrofitting required in older schools, including electricity and HVAC systems, which we estimated could affect up to 23% and 4% of schools respectively. These costs were estimated to be $240,000 for electricity and $31,800 for HVAC in an average school. Obviously, these costs will vary by age and condition of school, as indicated in the body of the report. A computer replacement cycle of 7 years and 5 to 10 year replacement cycles for the other equipment were incorporated into the ongoing operations and maintenance costs.
We assumed multimedia-capable computer prices of $1,700, a typical price paid today by K-12 schools. We further assumed that this price declines by 4% per year. This relatively small price decline is based on the assumption that schools will continue to purchase multimedia-capable computers that have enhanced functionality as it becomes available and that provide special access features for physically impaired students (e.g., written instructions for the hearing impaired, sound for the sight impaired, and special manipulatives for the physically challenged). This viewpoint is validated by the historical trend and is shared by a number of the major hardware manufacturers, who have plans to add functionality and believe that consumers-including those in the schools-will value the upgraded capabilities for at least the timeframe we consider here.
In addition to each computer, we assumed 2 printers ($535 each) and scanners ($675 each) for the Lab model, and 1 printer and scanner per classroom for the Classroom model. Furniture and security equipment were also included ($355 per computer and $350 per room).
We estimated 14 multimedia-capable computers per school today based on installed base statistics and 1994-1995 shipments. (See Exhibit 18: "Instructional Multimedia Computers Per School.") However, these computers are distributed unevenly across schools. We have taken this uneven distribution into account in the Lab model; the adjustment represents approximately a 10% increase in hardware costs. In addition, we assumed an installed base of 1 printer and 3 security/furniture stations per school.
d. Content
Content costs include prepackaged software and access and usage charges for on-line services. Software upgrades were assumed to be annual or biannual depending on the particular package or service. Ongoing assumptions for software included expenditures for bilingual capability where applicable. While we made specific assumptions about prepackaged software versus services, our belief is that these costs are interchangeable. In total, the expenditure on software for the Lab model in the year 2000 is 30% higher than expenditures on all electronic media today; for the Classroom model, the expenditure in 2005 is 230% higher than today. Future costs were assumed to decrease at 3% per year.
According to NCES data, approximately 35% of schools currently have access to the Internet or commercial on-line services. Once again, however, most of these connections are available only in the school library and/or media center.
e. Professional Development
These costs include substitute teachers (at $100 per day) to cover times when teachers are out for training, as well as support resources-1/4 full-time equivalent (FTE) in the Lab model and 11/2 FTE in the Classroom model-shared across the district to help teachers integrate the technology into the curriculum. Costs for the training courses themselves were also included.
In concert with Teaching Matters, we estimated that 50% of the teachers are at the entry level, 25% at adoption, 20% at adaptation, and 5% at appropriation per thestages shown in Exhibit 14: "Teacher Skill Stages." In the Lab model, trainees (teachers, school administrators, librarians, and selected district personnel) receive sufficient instruction to attain basic adoption level (30 hours); in the Classroom model, 80% of teachers are trained to the adaptation level and 20% are trained to a higher level.
f. Systems Operation
Systems operation costs include resources shared across the district dedicated to designing and operating the systems. The initial deployment costs for the Lab and Classroom models are $5,300 for design charges and 1/4 FTE and 1/2 FTE respectively. These same FTEs are assumed on an ongoing basis.
Video Infrastructure
Two video infrastructure models were costed: a business-quality video facility and a low-end professional-quality video facility. These models were costed as incremental to the Classroom (or Partial Classroom) model. Both models assumed a single video room with a monitor, three cameras, soundproofing material, and microphones. The business-quality facility has a T-1 connection and assumes equipment at a price of approximately $19,000. For 50% of rural schools, we assumed wireless radio with a POTS backchannel (instead of a T-1 connection). The low-end professional-quality facility has a T-3 connection and assumes equipment at a price of approximately $46,000. Telecom charges were based on average RBOC tariffs.
In addition, initial professional development costs were assumed to be $1,775 per school for teachers, and initial and ongoing system operation costs were assumed to be $9,300 and $11,240, respectively, representing a part-time facilitator/system administrator.
Voice Infrastructure
Costs were also estimated for providing voice mail to all schools and for placing telephones in all classrooms. The voice mail costs are independent of the computer-based models, but the classroom telephones assume that classroom wiring is already in place (i.e., the Partial Classroom or Classroom models).
The voice mail option assumes a dedicated voice mail server for each school ($1,500) and the use of one POTS line. Costs for initial training were assumed to be $1,000. No additional allowance was made for ongoing support; it was assumed this would be handled by dedicated computer support staff.
For the classroom telephone option, 1 telephone per classroom was assumed with 4 telephones per outside line; schools install multiple new POTS lines connected to a concentrator. Once again, costs for professional development and ongoing operations support were assumed to be minimal. The national costs for video and voice infrastructure, by the six elements, are displayed in Exhibit 19.
|
|
Footnotes
55 Averages for 1994-1995 included: 5.7 schools per district, 533 students per school, 31 teachers per school, 21 classrooms per school, and 25 students per classroom. These averages are derived from figures provided by the National Center for Education Statistics (NCES).
56 We utilized the following numbers from the National Center for Education Statistics: 84,500 schools, 14,850 districts, 45.0 million enrolled students, 2.6 million teachers, and 1.8 million instructional rooms. The student population is expected to grow by 7% in 2000 over the 1995 base and by 10% in 2005, according to the Department of Education.
57 Fixed wireless solutions have a number of limitations, particularly in urban or suburban environments: a clear line of sight is required, reliability can be low, only data and digitized video can be transmitted, and there is potential for clogging the bandwidth as more and more users seek to utilize wireless communications.
58 See supra note 26.
59 2 mbps wireless LANs have been in existence for some time and proven reliable; 10 mbps LANs (Ethernet equivalent) have recently been introduced and early trials are promising. While their relative price makes wireless LANs attractive wherever remediation would be required, many school buildings have structural barriers that make their use impractical.
60 For further discussion on this point, see supra note 28.