The pace of deployment for any of these infrastructure models depends on three factors: funding availability, professional development, and courseware availability. Schools' ability to acquire computer equipment and network their facilities is primarily a matter of obtaining funding, but the value of the hardware and network connections depends on the quality of the applications and teachers' ability to integrate them into the curriculum. In other words, simply raising the money for the physical infrastructure is not enough: teachers, courseware developers, and community leaders must come together if the benefits of the infrastructure are to be realized.
Consequently, deployment presents a number of challenges for schools. First, districts need to raise funds for installation and ongoing operations in the face of competing demands for funding and budget cutbacks. Second, teachers need both incentives and time to develop the new skills required to make effective use of network technology through both formal training and hands-on experience in the classroom. Third, a wide selection of high-quality multimedia courseware needs to be made available to supplement the traditional textbook-based curriculum.
These challenges increase as schools progress from the relatively simple goal of promoting computer literacy to more ambitious efforts to use network technology as an integral part of the curriculum. While a lab may be sufficient for basic computer-based assignments, networked computers need to be in the classroom if they are to be used as part of the day-to-day learning experience. Broad deployment, in turn, raises the funding hurdle and puts much greater demands on teachers. A broader selection of courseware is also required to meet the needs of a wide range of subjects and grades.
Although these challenges are substantial, they are surmountable. Funding needs can be met by a combination of reducing costs, reprogramming existing educational funds, and obtaining funds from new sources. Teachers' skills will develop with appropriate incentives, on-the-job experience, and in-service training; revised certification requirements and teacher college curriculums will also help reinforce this goal. Finally, the courseware market will develop as demand mounts from schools that have deployed the infrastructure and teachers search for new on-line content.
Meeting the funding challenge
The funding challenge is substantial both because of the limited access most schools have today to the basic infrastructure, and because of the fiscal pressures at work in the current budgetary environment. Setting budget priorities among many competing demands for funds-and securing grants, donations, and subsidies-requires strong leadership at many levels and a clear, compelling vision, as well as a good dose of creativity and persistence.
Limited current infrastructure. When it comes to basic infrastructure, most schools are starting from a low base. While many schools have computers, as of 1994 over 85% of these computers were not equipped to support the latest multimedia courseware-in other words, they could not combine text with advanced graphics, video or sound. Neither could many connect to an internal or external network. Factoring in new computer purchases in the 1994-95 school year, there are now on average 14 multimedia-capable computers per K-12 school or approximately 38 students per multimedia-capable computer. However, averages are misleading: the computers are not evenly distributed across schools. Surveys conducted by Quality Education Data, Inc., reveal disparities across schools based on socioeconomic and racial/ethnic status, although the situation has been corrected to some extent through federal funds and special grants available to underprivileged areas. For example, public K-12 schools with less than 20% of students qualifying for Chapter 1 funds (i.e., students from low income families) average nearly 8.6 computers (of any type) per 100 students while schools with over 80% average only 7.2 computers per 100 students. Likewise, schools with no minority students average 9.9 computers per 100 students while schools with over 90 percent minority students average only 7.3 computers per 100 students. (27)
Similarly, the external and internal network connections in schools today are limited. While 49% of schools have local area networks, half of those connect administrative computers. Fewer than 10% of these networks connected computers in all classrooms as of the 1993-1994 school year. (28) Likewise, although most schools have telephone lines, almost all are for administrative use; only 12% of classrooms have telephones. (29) Fewer than 5% of schools have high-speed, high-quality ISDN or T-1 connections, (30) and rough estimates from telephone companies indicate that up to one-third of schools are in areas where ISDN and T-1 connections are currently not available. Furthermore, while over 70% of schools have cable installed and up to 35% have satellite hook-ups, little of this infrastructure is currently capable of handling interactive applications. (31)
Budget pressures. To place the funding discussion in context, about 1.3% of the national public school budget is currently spent on instructional technology.(32) As discussed above, current spending would almost cover nationwide deployment of the Lab model, which would consume at most 1.5% of the nation's annual education budget. (This is a nationwide average; as mentioned above, the percentage of an individual school's budget going to technology would vary.) The Classroom model, on the other hand, poses a much greater challenge: the instructional technology budget would need to triple to meet the 3.9% of spending that this model would require. However, a continuation through 2005 of the recent technology spending growth rate of 16.5% per year would come close to reaching that 3.9 level-if this growth rate can be sustained. (See Exhibit 9: "Projected School Instructional Technology Spending.")
Sustaining such a growth rate, however, will not be easy. The education budget is caught between upward pressures on spending due to demographics, inflation, and other demands, and downward fiscal pressures on government spending programs. The $249 billion per year that is currently spent overall on public K-12 education is forecasted to grow at a rate of 5.6% per year through 2005. About 1% of this increase comes from predicted growth in the number of students, and 3% from inflation, leaving only 1.6% for all other increases in per-student spending.(33) And given mounting pressures for cuts in federal, state and local budgets, this projected 5.6% growth rate may not materialize, further constraining technology spending.
Other important demands for educational funds also will compete with technology for share of the budget. Basic repairs and facilities upgrades (estimated at $101 billion) are a top priority for many schools, as are school security programs. (34) Mandated programs, such as compliance with federal requirements for asbestos removal and handicapped access ($11 billion over the next 3 years) are also contributing to budget pressures. (35) Finally, teachers' salaries-currently 57% of educational spending-have increased faster than inflation over the past decade. (36) Technology requires funding not just for the initial installation, but also for ongoing operations, training, upgrade and maintenance costs. Locking sufficient funds into the budget over the long term implies that these budget battles will need to be fought year after year.
Despite these budgetary pressures, our analysis suggests that the funding challenge can be met through a combination of cost reduction, reprogramming existing funds, and additional initiatives from both private and public sectors. For example, the Classroom model could be funded by the following combination of initiatives: maintaining the current spending rate on technology of 1.3%, capturing 0.4% through additional cost reductions (or a further 10% savings on purchases), reprogramming anywhere from 1% to 2% of closely related budget categories, and securing up to 1% in additional funds. The more successful the cost reduction and reprogramming initiatives are, the lighter the burden that will fall on securing alternative funds. The following list of funding suggestions is neither prescriptive nor by any means exhaustive.
Reduce costs. One way to reduce the cost of deployment is to form buying consortiums at the state, regional, or national level to negotiate lower prices than a typical district could negotiate on its own. Such negotiation with equipment and service providers could reduce the cost of deploying the Classroom model by about 10%; these savings go beyond discounts assumed in the model. (See Exhibit 10: "Estimated Potential from Cost Reductions.") Likewise, securing donations of in-kind services from local community groups-free local area network installation, for example-represents another way to reduce individual schools' funding burden.
Cost reduction efforts should target the largest cost elements that can be affected: hardware, internal network installation, and professional development for teachers. Most proposals to date, however, have focused on the connection to the school-for example, ensuring universal access to the Internet through telephone line or other connections. While such initiatives are important, they will not by themselves make much of a dent in overall funding needs.
Reprogram existing funds. A second set of actions focuses on shifting existing educational funds to new uses. Selected categories of the school budget are natural candidates for potential reprogramming in support of connecting schools (see Exhibit 11: "Distribution of School Expenditures, 1992"). Textbooks account for about half of schools' expenditures on "instructional materials, supplies, and services"-about 2% of total school spending. Some of these funds could be used for multimedia courseware and on-line instructional materials, supplementing (or replacing) traditional textbook purchases. Another 8% of school spending is currently devoted to "instructional support," such as instructional supervisors (e.g., the head of the math department). Some of these resources could be redeployed to address teacher training and support needs. For example, instructional supervisors could focus on helping teachers integrate technology-based tools into the curriculum.
Reprogramming funds within these natural candidate categories could contribute 1% to 2% to the technology budget (see Exhibit 12, previous page: "Estimated Potential from Reprogramming Funds"). In addition to this 1% to 2% from natural candidates, some general funding categories can also be reprogrammed. In Carrollton, Georgia, for instance, the district cut administrative staff by 20% to 30%, releasing funds for technology and connection within their schools. Some schools, such as those in the Hueneme District, have chosen to fund technology rather that teachers' aides.
Secure additional funds. A third funding option-and perhaps the most difficult-is to secure new sources of funding. Currently, state and local government funds cover 84% of the public K-12 education budget, but account directly for only 60% of technology spending (see Exhibit 13: "Sources of Public School Funds"). Some state and local governments have issued special educational bonds, increased taxes, and/or allocated lottery funds to cover investment in educational technology. A range of other funding sources have provided support for technology to date, including federal Chapter 1 and 2 funds.
Innovative schools and districts have also found a number of ways to raise money from local community groups, private industry, and foundations. Some schools and districts have been fortunate enough to be chosen as model schools or pilot sites for major equipment suppliers including telephone, cable, and computer companies. Others have received special grants from a range of sources, including private foundations. Some have set up entrepreneurial ventures such as developing and selling their own educational software. The Carrollton School District offers one good example of a creative approach to funding. (See sidebar, "Case Study: Carrollton School District, Georgia.")
Providing professional development
As discussed above, the greatest benefit from connecting schools to the information superhighway is derived when the technology is fully integrated into the curriculum. Integration into the curriculum requires that teachers be able to use the technology effectively in whatever subject they are instructing. In turn, this requires professional development for the vast majority of teachers: first, to master the technology; second, to learn new teaching methods incorporating the technology. In addition, it requires professional development for those who advise and support teachers: school librarians, media specialists, and administrators.
According to Teaching Matters, a New York-based education consulting firm,(37) almost 50% of teachers have little or no experience with the relevant technology. The current system does little to support teachers in acquiring these skills. Few teachers have full-time access to a computer at school. In addition, there is little opportunity or incentive to gain pre-service training in technology. Only 18 states include any technology skills among the requirements for teacher credentialing. And in most states, the requirements are too low to matter. Most schools and colleges of education have relatively little technical equipment or resources, and they devote limited course time to preparing teachers to use technology effectively in the classroom. In California, for example, a year-long teacher training program includes a total of only 11 hours instruction on the use of computers and little or no instruction on networking or other aspects of the NII.
In-service professional development opportunities offered or required by schools and districts vary widely, but generally tend to be minimal. QED reports that 81% of school districts spend less than 10% of their technology budgets on training.(38) Based on a survey of its readers-who are likely to be relatively sophisticated technology users-Electronic Learning reports that only 8% of technology budgets went to training.(39) While these numbers are likely to understate training and support, they are consistent with our own case studies and interviews, which indicate that even most "model" technology schools spend no more than 15% on training and support. By contrast, the "Teacher Skill Stages" model (to be discussed below) calls for substantially greater expenditures. Experience in the corporate sector shows that investment in training is crucial to getting the benefits out of technology. Likewise, teachers, as well as corporate employees, need to learn both how to use the technology and how to do their jobs differently.
Most of the in-service training in technology skills that teachers do receive is at best exposure rather than real skill building. Electronic Learning found that only 21% of training courses are geared toward integrating technology into the curriculum.40 In addition, half of all training is delivered in the form of a half-day workshop.(41) A lecture or half-day seminar with little or no follow-up or in-classroom support is unlikely to promote either mastery of the technology or changes in teaching approaches to incorporate the technology.
Finally, teachers have little incentive to pursue aggressively the type of professional development needed to integrate technology into the curriculum. Districts that tie pay scales or recertification to continuing education rarely mandate a technology component. Furthermore, most college entrance requirements do not address technology competence or the use of technology in exploring academic subject areas. Consequently, there is little motivation for K-12 teachers to regard technology as playing an essential role in preparing their students for college.
Nonetheless, we have observed many teachers taking the initiative to learn and use computers and the NII in their teaching. And, of course, we found a few districts where teachers were given the lead role and the time and support needed to master the technology and integrate it into their curricula. "We gave six teachers a full year to think through the technology and connectivity they wanted, the physical layout of the classrooms, and the ways in which they integrated technology into their courses," explained Dr. Ron Rescigno, District Superintendent of the Hueneme School District. "They were encouraged to attend conferences and to network with other teachers and professionals associated with technology. Obviously, we still provide real-time support to our teachers and special technology courses on an ongoing basis, but the ability of these six teachers to completely focus on creating their own technology environments has made a huge difference. Talk to any one of those teachers or their students-they are delighted with the outcome."
Another model is offered by the Ontario-Montclair school district, the second largest K-8 district in California, where a large portion of the teacher training occurs in a computer training lab maintained at the district headquarters. Dick Archibald-Woodward, the technology coordinator for the district, believes strongly that teachers require instruction in how to integrate technology into the curriculum, as well as in developing basic computing skills.
The district computer training lab, which Archibald-Woodward manages, has about 30 computers, including Apple IIes, multimedia-capable Macs, and PCs. The instructors are 12 teacher/mentors who add this responsibility to their normal teaching load but are given a supplementary salary stipend. Approximately 400-500 teachers undergo some type of technology training in this center each year. The training programs cover a wide range of topics, including basic computing skills, specific applications, curriculum integration, and networking. Teachers can receive college credit for some of the courses. Some of the training is compensated through regularly scheduled teacher in-service days and release-time training. The district's training program also includes a number of training courses and seminars provided at school sites, as well as support staff who are available to visit sites and provide "just-in-time" training and support as needed.
An interesting twist on teacher professional development is made possible by the technology and connectivity itself. For example, the "Online Internet Institute" is a newly formed initiative that is leveraging the Internet to bring together a group of 665 educators from school districts around the country during the 1995-1996 school year. These educators receive instruction on-line about integrating the Internet within their classrooms and supporting their peers in doing the same. This instruction is provided by on-line mentors and includes access to information resources and support for curriculum integration (e.g., lesson plans, technology suggestions).(42)
In addition to in-service development opportunities, some states and colleges of education are taking the lead in establishing higher standards of competency with technology and providing the resources-including equipment, course time, and expertise-to ensure better preparation of teachers entering the school system. For example, the School of Education at Northwestern Kentucky University is actively working on increasing the requirements for technology training beyond a one or three semester hour course. The School is investing in new technology infrastructure both for its computer lab and in support of its computer-aided classes, in which each student is provided a computer and modem for the semester. Most of this activity anticipates the implementation of a state-wide technology plan for the K-12 public school system.(43)
In Texas, the Houston Consortium is focusing on completely redesigning teacher education. The Consortium's effort to integrate technology into the pre-service education of teachers is particularly significant. Each prospective teacher is encouraged to purchase a laptop computer for lesson planning, telecommunications, record keeping, and instruction. The Consortium also supplies each participating K-12 school a telecommunications center and a portable multimedia station to be used by the pre-service (and in-service) teachers. Finally, the Consortium also provides both individual laptop computers for the professional development of up to 6 faculty members and a computer classroom (5 computers for instruction and 10 laptops for students) to each participating university or college of education. Training is provided both in the use of the technology and in the integration of the technology into the curriculum. To date, the results have been extremely encouraging, and the organizers and participants continue to pursue multiple initiatives in order to make sure that "graduates from the programs of the participating colleges of education will enter the classroom as new teachers with knowledge and skills in the use of technology that will match their knowledge of subject matter and their skills in teaching children."(44) As examples like these suggest, no one model for teacher professional development will be right for all schools, districts, and states. However, we believe that some basic principles will help many schools get started and some broader actions could provide valuable support to local school and district initiatives.
A first step is to set accurate expectations as to how long effective professional development is likely to take. Exhibit 14, "Technology Skill Stages for Teachers," shows a five-stage professional development model based on our analysis, with input from Teaching Matters. Moving teachers from entry through the first two stages could be achieved in half a school year for any one teacher. The prerequisites are adequate access to a computer, courseware to enable the use of technology in the curriculum, and support for the teacher in the classroom. Ideally, the support would come both from experts in the technology and from peer teachers. This implies giving teachers time and encouragement to share experiences with each other.
Experience at schools that have been down this path suggests that the two more advanced stages on the professional development model simply take time-from two to five years of real teaching experience with the technology. In addition, progressing to these stages requires encouragement and incentives for teachers to make the extra effort needed to build their own skills and support other teachers. Thus, a school district that starts now with basic "Adoption" and "Adaptation" training could build a population of appropriately skilled teachers over a six- to seven-year period (assuming two years to move all teachers through the basic training-an aggressive assumption, to be sure).
In the meantime, we believe several actions are appropriate for most schools and districts to consider:
An aggressive professional development effort involving the support of teachers, administrators, boards of education, states, the federal government, and schools of education will be an essential part of effectively connecting students to the NII.
Ensuring courseware availability
Today, the market for courseware is relatively small, fragmented, expensive to enter, and risky. As a consequence, it is underdeveloped-although this will change as K-12 school demand for courseware grows.
For purposes of this discussion, we have defined courseware as "electronic curricular materials." Courseware includes interactive multimedia software, on-line educational services, teacher's guides, and other materials linked directly to prescribed curriculum. The link to curriculum is critical because teachers have a limited time to cover concepts and facts outlined in the curriculum. Good courseware allows students to work in groups and at their own pace, and to receive quick feedback on their progress.
For production of high-quality courseware to flourish, the courseware market needs to expand and to become more attractive and accessible both to existing and to new providers. Fortunately, as more schools commit to connecting to the information superhighway and find the funding to do so, and as more teachers become knowledgeable and excited about using technology in their classes, demand for courseware will naturally grow. Even so, it might be worthwhile to consider options for stimulating growth in the courseware industry-for example, speeding up the schools' slow and bureaucratic procurement processes-to make sure that enough good courseware is available to encourage schools and teachers to experiment with technology in the near term. Small,fragmented market.
Just a piece of the overall education market, courseware comes in two basic types: (1) integrated programs that typically support a full-year course, and (2) more tightly focused, modular programs that cover a specific topic (e.g., the Oregon Trail, the writing of the Constitution). The market for both types of courseware totaled about $290 million in 1993-1994.(47)
At $290 million, the courseware market is smaller than other software markets. One particularly relevant comparison is to the home market for educational applications, since developers who have chosen to focus on the education market have told us that the home market is most attractive. LINK Resources estimates the size of the home education software market at $1.4 billion in 1995, and the home "edutainment" market at nearly $500 million.(48) Not only does this substantially exceed the size of the K-12 school market, but it is expected to grow at a more rapid pace over the next several years. The growth in the home market is supported by the increasing penetration of multimedia computers into the home. The number of multimedia comput-ers used for instruction in K-12 schools is projected to grow from about 1.0 million in 1994-1995 to 2.2 million by the 1997-1998 school year,(49) while the number in the home is forecasted to grow from 8.0 million in 1994 to 38.3 million in 1997.(50) If these projections hold, then the number of multimedia computers at home will exceed the number in schools by a factor of 24 to 1 by 1997.
The size of the courseware market is further constrained by the distinction made between core and supplemental materials. By rule, state textbook monies typically go to core materials. Because courseware is normally considered supplemental, this reduces the available pool of dollars for courseware purchase.In addition to its relatively small size, the courseware market is fragmented into numerous small segments. Programs need to be tailored to different academic subjects and to individual grade and skill levels. While multimedia courseware lends itself to interdisciplinary content that could combine subjects, state curricula are not currently written in a fashion that would lead to approval of most courseware for multiple subject areas.The combination of a small market, fragmentation, and a relatively more attractive home market has created a chicken- or-the-egg dilemma for courseware developers. If the demand for courseware were larger, developers would produce more and better educational products. On the other hand, the limited spectrum of available products inhibits the development of infrastructure and therefore the growth of demand.
High cost to serve.
The developers we interviewed regard the educational market as a difficult place to do business because sales and service are complicated and expensive. Schools' purchasing process is slow and arduous. Approvals are required at many levels and each decision maker has a high need for information.
Twenty-two states select course materials through an "adoption" process that poses three hurdles for courseware developers. First, the interval between selection of materials for a given subject and grade is long-often five years or more. While this may be appropriate for textbooks, for which the process was designed, it is less desirable for software, which changes rapidly. Second, the sales process is expensive and risky, particularly for smaller developers. For example, the textbook choices of Texas and California carry significant weight throughout the country. As a result, vendors spend heavily-with no guarantee of success-to lobby the committees of teachers and other stakeholders who recommend materials in these states. After participating in the adoption process in one of these major states, one developer of highly acclaimed courseware said that it could not afford to do so again for many years. Third, the sales cycle does not necessarily end with adoption. In states that select more than one text, adoption merely signals that the next phase of the sales cycle has begun, this one directed to district- and school-level officials.
In addition to the difficulties with the adoption process, the mechanics of school district purchasing practices are often cumbersome. District agents require purchase orders tailored to their own unique systems. Some want to be billed after the goods have been received; others before. Some are restricted from paying until the product has been fully consumed, which is particularly difficult for a product that is part software and part on-line service. Others put off buying until the end of the budget cycle, ordering if they have money left over and requiring delivery within the week. The combined effect of such procurement practices is to raise the costs providers must bear.
The schoo courseware market is also costly to service due to high training needs. Pioneer providers often face high training costs because teachers are simply not familiar with computers and networks. One developer of a networked application stated that by far the main reason for calls to its help line was that the teacher did not understand how to connect to the network.Risks of product development. Courseware is relatively expensive to develop and comes with little guarantee of success. The experience of multimedia developers generally is a good illustration of the risks faced by courseware developers specifically. A survey of 912 multimedia software developers conducted by Gistics, a California consulting firm, concluded that 96% were unprofitable.(51)
In addition, multiple platforms further increase the costs of production. The public schools have a mix of Apple Macintoshes, IBM-compatible computers, and older Apple IIe and Commodore machines. While new applications and developers generally aim at the new machines, porting an application developed for the Apple Macintosh Operating System to the Windows operating system can add 10- 20% to its cost.
Addressing the courseware challenges. As mentioned above, some of these problems are likely to sort themselves out over time as more schools begin using computers and networks in the classroom, and the market for courseware grows as a result. However, there are steps that could be taken now to stimulate the courseware market in the near term. It is hard to know just how important such steps would be, but they seem to be worth careful consideration.
Perhaps most important, there are a number of ways to address the small size of the courseware market. Clearly stated national goals for deploying technology in the schools, state technology plans, and real appropriations could build confidence among courseware providers that demand will grow and that the growth will be sustained. In addition, changing the rule in many states that prevents textbook money from being spent on courseware would help. Twenty-one of twenty-two adoption states have taken steps in this direction by redefining instructional materials to include electronic content. The next step would be to relax the distinction between core and supplemental materials.
Furthermore, the fragmentation of the market into small segments defined by grade and subject is not inevitable. Instead, wider skill-based, cross-disciplinary segments could evolve. Many districts and states are systematically rethinking and updating their curricula. To the extent that the new curricula emphasize flexibility of method and skills over content, this would encourage the formation of these larger, more profitable market segments.
The high cost to serve the K-12 market can also be addressed. Districts can streamline their purchasing practices. Friendlier adoption rules for courseware can be created. And training and support at the school level can be enhanced so that early developers do not have to bear the brunt of training teachers in computer basics and solving their particular hardware problems.
To mitigate the risks faced by early developers, states and districts can enter into partnerships with developers. Agreements might range from providing venture capital, to cooperative development arrangements, and to advance agreements to purchase. For instance, the state of Florida has established a fund to encourage the development of courseware that meets its curriculum needs. In return for providing seed funding, schools within the state receive a discount on packages purchased. Money earned by the state on its investment is returned to the fund, which has just seen its first product complete the cycle through development to sales to dividends. When the Guilford County School District in North Carolina wanted teacher productivity and student performance management software, it scoured the market but could not find the product that met its needs. So it contracted with McGraw-Hill to build the system; McGraw-Hill was pleased by the deal because it reduced the risks of development.
Grants have also been used to stimulate the development of high-quality courseware. Several challenge grants from the National Science Foundation (NSF) have been focused on courseware or the underlying tools to create it.52 For instance, The Geometer's Sketchpad allows students to test hypotheses in real time on geometric models they create on the computer. Students can explore the model by manipulating objects and observing how the other objects respond. Students' observations can be visual, or they can measure the resulting angles, lengths, and areas using tools built into the program. The Sketchpad grew out of the Geometry Forum, a project at Swarthmore University funded by the NSF.(53) Footnotes
27 Technology in Public Schools: QED's 13th Annual Census of Public School Technology Use (Denver, Colorado: Quality Education Data, Inc., 1994), pp. 26-27. 28 Market Data Retrieval reports that, during the 1992-1993 school year, 49% of schools had a local area network for any use; see K-12 Education Market Report (Washington, D.C.: Software Publishers Association, July 1994), p. 31. QED reports for the 1993-1994 school year that 23% of schools had a network for instructional use, of which 18% (or 4% of all schools) connected classrooms; see Technology in Public Schools, supra note 27, pp. 76-77; see also Educational Technology Trends, QED's 7th Annual Sample Survey of Technology Use and Purchase Plans in U.S. Public Schools (Denver, Colorado: Quality Education Data, Inc., 1994), p. 56. 29 Princeton Survey Research Associates, "National Education Association Communications Survey: Report of the Findings" (Washington, D.C.: National Education Association, 1993), p. 2. 30 National Center for Education Statistics (NCES), Advanced Telecommunications in U.S. Public Schools, K-12 (Washington, D.C.: U.S. Department of Education, Office of Educational Research and Improvement, February 1995), p. 13. Of the 49% of schools reporting wide-area network access, 3% report having a T-1 connection, and 4% an ISDN connection, suggesting that 3.5% of all schools have access to either one. In addition, 4% of the 49% reported access to "other" connections. 31 Ibid., p. 7; Margaret Honey and Andrs Henriquez, Telecommunications and K-12 Educators: Findings from a National Survey (New York: Center for Technology in Education, Bank Street College of Education, 1993), p. 11. 32 See Appendix C for the breakdown and derivation of this figure. 33 National Center for Education Statistics, Projections of Education Statistics to 2005 (Washington, D.C.: U.S. Department of Education, Office of Educational Research and Improvement, January 1995), p. 83. 34 Not all repair and upgrade expenditures are inconsistent with technology spending, however. In fact, retrofitting schools to accommodate technology can be effectively coordinated with some repairs and upgrades. See Ezra D. Ehrenkrantz, "Retrofitting in Increments: Redesigning Your School for Whatever the Future May Bring," Electronic Learning (February 1995), pp. 22-23. 35 U.S. General Accounting Office, School Facilities: Condition of America's Schools (Washington, D.C., February 1995), pp. 5-7. 36 From 1980 to 1993, teacher pay increased relative to inflation (20% higher). "Will Schools Ever Get Better?" Business Week, April 17, 1995. 37 Teaching Matters, Inc., a not-for-profit organization, has worked extensively with K-12 schools in and around New York City to develop, monitor, and deliver professional development programs (formal training plus ongoing support) which help teachers and principals to integrate technology into their classrooms. 38 Educational Technology Trends, supra note 28, p. 11. 39 Jessica Siegel, "The State of Teacher Training: The Results of the First National Survey of Technology Staff Development in Schools," Electronic Learning (May/June 1995), p. 44. 40 Ibid. 41 Ibid., p. 48. 42 Interviews with Bonnie Bracey, cofounder of the Institute, September 1995. 43 Connie Carroll Widmer and Valeria Amburgey, "Meeting Technology Guidelines for Teacher Preparation," Journal of Computing in Teacher Education, vol. 10, no. 2, pp. 12-17. 44 Richard Alan Smith, W. Robert Houson, and Bernard Robin, "Preparing Preservice Teachers to Use Technology in the Classroom," The Computing Teacher (December/January 1994-1995), p. 59. 45 Consortium for Policy Research in Education, CPRE Policy Briefs (June 16, 1995). 46 Ibid. 47 This estimate of the size of the courseware market is based on several sources. As mentioned above, there are two types of courseware; the types are called integrated learning systems, or ILS, and modular, unit-based software. An ILS is a turnkey package that typically supports a full-year course, comes packaged with student management and testing tools, and sometimes includes hardware. Despite their breadth, ILSs are still considered supplemental to textbooks because they typically lack the depth necessary to completely cover a full-year core curriculum. The Software Publisher's Association estimates the software portion of the ILS market at $170 million for 1993-1994. ILSs of the past often had features which caused them to fall out of favor: proprietary hardware, software that did not work with other packages, and a drill-and-practice orientation. ILSs have given way to what one analyst has termed "networked learning systems." By contrast, modular, unit-based software focuses on a single topic or concept. The size of the market for unit-based software is not tracked separately from the $360 million that schools spend on non-ILS software, which includes edutainment, reference and on-line software and services. Based on interviews and case studies, we estimate that unit-based software accounts for about one-third of this total, or $120 million. For information about market size, see K-12 Education Market Report, supra note 28. For information on market definition, see the Smith, Barney report on Davidson & Associates, August 3, 1993. 48 Consumer PC Market Outlook: 1994-1999 (LINK Resources Corporation, June 1995), Tables 6 & 9. The Software Publisher's Association estimates the size of this market for 1994-1995 at $630 million. Edutainment software combines education with entertainment, often in the form of multimedia games. Edutainment products are typically not curriculum-linked and their educational value varies widely. 49 The 1997-1998 estimate for multimedia-capable computers assumes that K-12 computer shipments continue to grow at 16% per year. 50 Consumer PC Market Outlook: 1994-1999, supra note 48, Table 4. 51 Jim Carlton, "Companies Aim to Dominate Fun Learning," The Wall Street Journal (August 2, 1995), p. B1. 52 Jerry Michalski, Release 1.0, Esther Dyson's Monthly Report (New York: EDVenture Holdings, Inc., May 1995), pp. 2 and 5. The report states: "The U.S. National Science Foundation (NSF) has funded many useful projects along these lines. In fact, almost every project we found intriguing was NSF-backed. It seems strange that NSF is the sole funder of so much activity. There's clearly a greater role possible for software developers and corporations." 53 Ibid., pp. 5-6.