Departmental Budgeting for Information Technology: A Life-cycle Approach Copyright 1994 CAUSE. From _CAUSE/EFFECT_ Volume 17, Number 2, Summer 1994. Permission to copy or disseminate all or part of this material is granted provided that the copies are not made or distributed for commercial advantage, the CAUSE copyright and its date appear, and notice is given that copying is by permission of CAUSE, the association for managing and using information resources in higher education. To disseminate otherwise, or to republish, requires written permission. For further information, contact Julia Rudy at CAUSE, 4840 Pearl East Circle, Suite 302E, Boulder, CO 80301 USA; 303-939-0308; e-mail: jrudy@CAUSE.colorado.edu DEPARTMENTAL BUDGETING FOR INFORMATION TECHNOLOGY: A LIFE-CYCLE APPROACH by John L. Oberlin ABSTRACT: Understanding the basic economics of information technology is a logical and necessary step toward resolving issues related to funding IT. Particularly in a distributed computing environment, IT planning and life-cycle budgeting can help campus and departmental administrators, faculty, and IT professionals make critical decisions regarding allocating increasingly limited institutional funds. This article is intended to assist deans, directors, department heads, information technology professionals, and other college and university planners to more effectively plan and budget proactively for information technology, especially local area networks. Although it covers many of the basic issues inherent in life-cycle budgeting for departmental technology, it does not cover the actual process of institution-wide strategic planning that is a critical basis for resource allocation. It is aimed primarily at larger institutions, although the model is readily adaptable to smaller college environments. The need for departmental planning and life-cycle budgeting Some administrators view information technology as a "black hole," where long-term planning is an oxymoron and budgeting is a never- ending stream of requests for "more." To others, the need for departmental planning and life-cycle budgeting--budgeting based on the practical financial "life" of an item--is unarguable, with the benefits being intuitively obvious. To still others, the problems of planning for information technology come not with the recognition of the need for life-cycle planning, but with the wherewithal (budget) to effect life- cycle planning. To them the lack of a current budget makes the exercise moot. Depending on the circumstances, each of these perspectives has validity. Addressing the issues embedded in this thinking is a challenge with which this article attempts to deal. It is based on the belief that understanding the basic economics of information technology (IT) is a logical and necessary step toward any comprehensive resolution of the issues. College and university administrators are increasingly asked to fund campus networks, upgrade existing network capabilities, build academic and administrative computing centers, add public and personal workstations, and establish departmental information systems. Yet, in only a few cases are these requests the result of a coordinated effort among all units affected. Moreover, the financial assumptions used to develop these budget requests are at times based more on political correctness than financial reality, thus contributing to the perception that IT is a black hole that can drain budgets with ever increasing speed. Life-cycle planning is intended to provide a seed for coordination, as well as information and education to faculty, departments, and campus administrators. Indeed, the assertion that administrators are much more likely to fund IT requests where the academic benefits and life-cycle costs are clearly stated is unassailable. In other words, IT planning and life-cycle budgeting are important tools that assist administrators, faculty, and IT professionals when making critical decisions regarding the allocation of increasingly limited institutional funds. Understanding technology life cycles Building and managing information technology systems is complicated by the extreme rate of change inherent in the industry. Thus, a key to good planning is to plan for change. One way to accomplish this is to understand life cycles and then budget accordingly. For the purpose of this discussion, "life cycle" is defined as the useful financial life of an item. In other words, the life cycle is the number of years you should plan to keep a piece of hardware or software. For example, a life cycle of three years for a computer implies that at the end of three years, the computer is either: (1) no longer suited for its intended purpose (e.g., Intel 80286-based servers won't run NetWare 3.11, currently one of the most important and popular network systems), or (2) maintenance and support have grown to the extent that it is cheaper to replace the computer than keep it, or (3) new requirements or performance standards (such as portability, ease of use, user interfaces, visualization, networking, processing power) have necessitated its replacement to meet user needs. A simple, yet practical example of deriving a life cycle results from thinking about the successive "generations" of advancement that most technologies go through. Consider microprocessors and the resulting desktop microcomputers. It is arguable that in the 80s, each generation of Intel processor was released approximately three years apart. Assuming a department or school needs to stay current with this technology, or could be no more than one generation behind, it would have to replace its microcomputers at least every four to six years--a four-to-six-year life cycle. Life cycles are not static, but vary depending on the technology and the rate of change inherent in the industry. For instance, in the 90s it appears likely that each new generation of microprocessor will be closer to two years apart, yielding a useful life cycle of two to four years, depending on its use and the nature of the unit. Departments and schools committed to leading-edge research may have more demanding expectations, while others may be satisfied with staying two or three generations behind. Advantages of planning and budgeting Both planning and budgeting for technology are essential. Together they help clarify goals, make assumptions explicit, and build consensus. In the simplest terms, life-cycle planning allows for a common vocabulary and perspective for the planning and sizing of information technology investments across campus. Other resulting benefits include: * a better understanding of the true costs of technology and support; * more coordination between departments and central campus organizations; * explicit financial delineations between central and departmental responsibilities; * more clearly defined expectations for technology and better strategies; * a more rational allocation of both departmental and central funds; and * better educated faculty, departments, and campus administrators. Central computing organizations Central computing organizations play a critical role in supporting departmental goals at their respective institutions. For example, at UNC- Chapel Hill, the Office of Information Technology provides core technologies in support of both productivity and leadership in the areas of collaboration, research, scholarship, and instruction. Additionally, OIT manages the campus-wide inter-building network that binds together much of the electronic communications of the campus. OIT's goal is to supply these core strategic information technology services with sufficient depth and breadth to allow academic units to leverage them cost effectively for their own ends. In this case, departments should not have to reinvent these services; however, they do need to make complementary investments if they want to leverage these central services. The line between central and departmental responsibilities can be fuzzy at times and is often based on legacy assumptions. Detailed IT budgets for both central and departmental investments can help clarify responsibilities and identify common needs. In this case it would be departmental life-cycle budgets for intra-building local area networks and central life-cycle budgets for the campus-wide inter-building backbone. The discussion that follows focuses on developing life-cycle budgets for the departmental or unit-level portion of the enterprise. Nevertheless, the methods described here can also be adapted to develop life-cycle budgets for larger and more complex central computing organizations. The need for local area networks As colleges and universities are increasingly asked to do more with less funding, the need to automate and computerize will continue to grow. Information technology is seen as one of the best hopes for improving productivity in research, scholarship, teaching, and administrative processes. Indeed, it is fair to argue that no institution will stay competitive in these areas without also keeping its investments in technology current. The future of information technology is increasingly tied to networking, access to remote information, and support of electronic collaboration. Budgeting for departmental local area networks (LANs) is an essential first step toward moving a department into the information technology arena. The value of financial planning While financial planning is an integral part of any comprehensive planning process, it is often overlooked. A financial plan, or assessment, is a necessary component of any larger plan that is intended to be acted on meaningfully. The life-cycle model described below is intended to be only a guide to help make the trade-offs inherent in technology life cycles more explicit. It does not cover any of the technical aspects that would be part of a comprehensive technology plan, nor does it cover the rationale for installing a LAN in the first place. Department and campus administrators should feel free to expand and modify this model to meet their special needs. For instance, it can be either a "back-of-the-envelope" baseline or reference point to compare current budgets and start a new planning process, or it can be a more rigorous tool used to form the financial basis of a strategic plan. Regardless of how rigorously life-cycle planning is adopted, it has several uses. Strategic planning. As mentioned earlier, any actionable plan will require that financial issues be addressed. From an implementation perspective, budgets are the link between plans and actions. They translate strategic plans into the financial resources necessary to implement the plan. Life-cycle planning is particularly useful when developing technology budgets, because it deals explicitly with the rapid rate of change inherent in the technology. Financial justification. The record of many institutions is replete with continuing requests for additional money for information technology. It would be naive to suggest that life-cycle planning alone could alleviate this. However, to the extent that it anticipates changes, translates expenses into steady-state budgets, and takes a comprehensive approach to identifying all the expenses over time that are necessary to support technology, it is a significant step in the right direction. Developing a budget that recognizes the inevitability of change and has a long-run perspective is a critical factor in breaking down the perception that IT is a budgetary black hole. Capital budgeting. For units with central IT responsibilities, life- cycle planning offers the opportunity to bring a more disciplined approach to capital budgeting. For example, consider the budgeting for public access microcomputer labs. Life-cycle planning not only sizes the problem from a very objective and long-term perspective, but also makes more explicit the financial and technological trade-offs of equipping a lab. Budgetary assessment. For departments that value information technology and are actively seeking to integrate it into the local infrastructure, building a life-cycle model is an excellent way to develop a financial baseline against which they can assess the level of current funding. If applied across departments or institutions, it can also be the basis for a broader assessment. Educating non-technical administrators. Many senior administrators are unaware of the dynamics, or basic economics, of information technology, and thus are poorly prepared to appreciate the financial demands inherent in managing an enterprise-wide information system. Life-cycle budgeting provides an easy-to-understand taxonomy that (1) highlights the differences between operating and capital budgets, and (2) explains the rationale from a simple, easy-to-follow, and common-sense perspective. Life-cycle categories As technology changes at an ever increasing rate, traditional strategies of planning for and financing technology investments will have to be reexamined. In any event, several generalizations can currently be made. In this case the assertion is that most technology investments can be assigned to one of five possible life-cycle categories. These categories are described below, from shortest to longest life cycle. Depending on the institutional environment and nature of the department, these life cycles might be extended up to 50 percent by carefully maintaining a homogeneous environment through coordination and standards. Careful attention to detail during the life-cycle budgeting process can make some of the trade-offs between diversity and lower cost more explicit. Personnel and support (annual expense). Personnel and support are annual expenses that are unavoidable costs of technology. Personnel includes staff for student labs as well as more skilled technicians who maintain local networks and file servers, and provide at least a minimum of on- site trouble shooting and training for the department. Support costs include annual maintenance and license fees, insurance, miscellaneous supplies, repair work for existing equipment, and training for both support personnel and users. Software (1-3 years). Software includes new software licenses and software upgrades running on all clients and servers, as well as operating systems and applications. Operating systems usually have longer lives than applications, but support considerations, especially in LAN environments, usually demand that they be kept current with new releases, typically every two years. Network software (e.g., NetWare) usually requires an annual license agreement and thus should be budgeted as a one-year life cycle. Application software ranges from one to three years in its usefulness. Although many applications can actually be used longer, network support issues and the need to share data usually require that applications be kept more current. Keeping software current is an increasingly important part of encouraging group productivity in networked environments. Hardware (2-5 years). Computing and network hardware (e.g., computers, printers, routers, bridges) usually have a life cycle of between two and five years. As the computing industry matures, the power and performance of PCs increase rapidly, while prices drop steadily. In fact, many industry observers describe the PC market as a commodity market, where product differentiation is declining and price competition is increasing. One result is that a PC for the average campus user may have a life cycle of about three years. This is merely an estimate for the "average" user. Many researchers will require shorter life cycles, while many others will have longer ones. As for students, it is becoming more and more clear that providing training on hardware that is over four or five years old may not be effective or institutionally competitive. Wiring (5-15 years). Network wiring, the copper or fiber in the walls, has a useful life of five to fifteen years. Network traffic and the demand for access to online information is increasing very rapidly. As usage grows, new technologies and protocols are emerging that allow some newer cable to be rejuvenated and transmit at increased rates. The usage trends are likely to continue indefinitely. How long the technology trends will continue is hard to estimate and makes predicting life cycles difficult. Most departments should expect current standards for copper wiring to meet their needs for ten years or more. Campus backbones will generally require fiber, and may have life cycles of twenty years. However, units that might want to run multimedia over the network today may require fiber to the desktop. Physical plant (30+ years). The information technology physical plant-- items like wiring closets and conduit--is a significant component of many new systems. Departments in 100-year-old buildings without conduit or dropped ceilings are often required to make expensive modifications before network cabling can be installed. In some cases, 90 percent of the initial cost of installing a network can be the one-time expense of making the building ready for cable to be pulled. Central computing units face the same issue when planning the campus-wide network. The possibility of encountering environments with asbestos, inadequate power supplies, insufficient cooling or space, or adopting uniform wiring standards, can also add significantly to the up-front cost of the basic physical plant for IT. Using a life-cycle budget worksheet The sample budget worksheet in Exhibit I has eight major sections for expenses and four columns to help calculate annual expenses. The annual expense is calculated by multiplying the number of units needed by the price per unit and dividing by the years of useful life cycle. We refer to this calculation as the life-cycle budgeting equation; it is the basis of all the budgeting calculations: (Number of units * price/unit / life-cycle years = annual cost) The calculation should be repeated for each line of the worksheet (except section 1) and then totaled at the bottom. Sections 6, 7, and 8 (maintenance, technical support, and contracted services) should be treated as annual costs and thus have life-cycle values of one year. Section 1, network wiring, should be treated as a one-time expense unless it is financed, in which case the annual payment should be used as the annual cost. The total annual cost is the cost of all equipment and support amortized over the expected economic life. This is the amount a department should expect to spend on average each year to purchase and support the technology. It is not, or does not have to be, the same as the actual cash flow. For units with a capital budget or other funding sources that can be considered annuities, the annual cost derived from the life-cycle equation can be mapped directly to the annual budget. However, units with more ad hoc funding for technology will have to use the annual cost as a baseline against which they allocate funds as they become available. Departments with both annual funds and ad hoc funds can use them in combination, perhaps using annual funds to fund the annual expenses (short life-cycle items) and using ad hoc funds in a more opportunistic fashion to fund longer life-cycle components. Regardless of the department's current financial condition, the life-cycle model provides a baseline against which it can assess its current condition and plan for future needs. The worksheet can also be used in reverse, by first starting with the annual funds actually available and then working the above equation backwards. In this fashion a department can work through a variety of "what if" scenarios, to see how changing the number, price, or life cycle will impact departmental plans. The major expense categories include the following: Network wiring. Network wiring is the intra-building wiring, office connections for both computers and peripherals, such as network printers or CD-ROMs, and related items necessary for the local area network. Also included in this section are the one-time expenses of installing conduit, wiring closets, and other physical improvements. Network hardware. Network hardware as used here means the hardware that is installed to manage the network or to provide services over the network. This hardware can be subcategorized in several possible ways, including: (1) servers for faculty, staff, or students; (2) routers and bridges (devices that allow LANs to communicate with other networks); (3) peripherals, including tape backup systems and CD-ROMs, and (4) other, various "black boxes" necessary for a fully functional network (e.g., dial-in modems, uninterruptable power supplies, dedicated phone lines, and network hubs). Desktop hardware. Desktop hardware is the equipment the end user has in the office. This includes the computer, monitor, network card, mouse, and other hardware as required. As part of the planning process, this section may be subdivided by users, so that the budgeting process can take into account their different computing needs. The numbers, prices, and life cycle may vary depending on the type of user. Printers. Printers are analogous to desktop hardware, except that they can be shared by more than one user. Thus the number is typically less than the number of computers. They can be subcategorized in various ways, depending on the size of the department or school. Software. Software includes all software installed either on the network servers or desktop machines. Subcategorizing software by either type of user, type of software (e.g., application vs. operating systems), or both, can help make the overall budget more accurate. For each subcategory the numbers, prices, and life cycle may be different. Network software is increasingly important and often overlooked (e.g., backup, e-mail, anti-virus, license control, etc.). Maintenance. Maintenance includes all annual service agreements and non- service agreement repairs for either hardware or software. In some cases, it may be more useful to do away with this category and include its subcategories with one of the areas above. Regardless of how the worksheet is organized, if service contracts exist, they need to be included in the budget as an annual expense. Technical support. Technical support is an often overlooked aspect of managing a departmental network. There are at least three basic aspects to support: personnel, training for support staff and users, and supplies. If a department is installing a LAN for the first time, or significantly expanding one, it should consult carefully to ensure that support costs are fully understood and budgeted. Contracted services. Contracted services are the computing and support services that are either better provided by a central computing organization, or are too expensive to implement and manage locally. This includes: (1) large-scale computing; (2) data storage, backup, and archive services, and (3) advanced technical support that only specialists can supply (e.g., microcomputer technicians, network consulting, etc.). ************************************************************************ Figure 1: Sample life-cycle budget worksheet EXPENSES NUMBER PRICE LIFE-CYCLE ANNUAL COST ________________________________________________________________________ 1 Network Wiring: Wiring Conduit 2 Network Hardware: Servers-Faculty and Staff Servers-Student Labs Routers and Bridges 3 Desktop Hardware: Faculty Staff Student Labs 4 Printer: Faculty and Staff Student Labs 5 Software: Faculty Staff Student Labs Networking 6 Maintenance: Clients Servers Printers 7 Technical Support: Personnel Training Supplies Connectivity Fees 8 Contracted Services: ________________________________________________________________________ TOTAL ************************************************************************ Note: This life-cycle budget was prepared by the UNC-Chapel Hill Office for Information Technology to help determine a student technology fee for student labs. The "percent of use" column was added to estimate what percent of that expense was allocated to students. Although OIT is optimistic about the long-run viability of this budget, it will likely be affected by the forces of "expectation inflation," "life-cycle optimism," and "investment creep" (see discussion of these reality checks later in this article). The life-cycle assumptions supporting this model will be reviewed and updated annually to reflect new understandings and realities. Operating and capital budgets In the simplest terms, operating budgets represent the costs associated with maintaining the status quo, while capital budgets represent the costs of replacing and growing information technology systems over time. With regard to financial planning, the critical issue is to realize that both are necessary, and to have one without the other will likely be untenable. Understanding the distinction between operating and capital budgeting is important, something that department chairs, deans, and central administrators will increasingly have to recognize and deal with. Operating budgets represent the unavoidable annual costs that are necessary to maintain existing systems. Capital costs are the unavoidable and necessary expenses that accrue over time as a result of product life cycles and the evolutionary nature of information technology systems. Failing to adequately plan and budget for the capital costs that accrue over time is often a major factor that contributes to the ad hoc nature of many information technology investments. Currently, central computing organizations characteristically have dedicated technology budgets, although many lack viable capital budgets that are distinct from their operating budgets. Departments are typically worse off; very few have an explicit information technology budget, and fewer still have any delineation between operating budgets and capital budgets. Figure 2: Sample life-cycle budget [FIGURE 2 NOT AVAILABLE IN ASCII TEXT VERSION] The exact delineation, or allocation, of expenses between operating or capital budgets is relatively unimportant. When segregation of capital and operating budgets is necessary, it is possible to use the life-cycle categories already discussed to allocate costs between operating budgets (by using short life-cycle categories) and capital budgets (by using longer life-cycle categories). Because capital budgets result from expenses that accrue over several years, it is desirable, if possible, that money allocated to this purpose be unencumbered by fiscal year boundaries. Developing capital budgets, and the attendant financial flexibility, is a major challenge facing administrators at all levels. Life-cycle planning considerations When attempting to predict and plan for the future, no process or method is perfect. Nevertheless, the life-cycle budgeting process is well suited for making explicit many of the critical assumptions that underpin the planning process. The following thoughts are included to highlight a few of the more important issues embedded in many financial plans. The policy issues inherent in these notions should be explicitly dealt with during the departmental planning process. * Although price/performance ratios are dropping, it is unlikely the total cost of information technology will decline significantly. Instead, the average annual expenditure for hardware is likely to remain constant or decline only slightly, as users, departments, and institutions tend to value greater performance over cost reduction. This implicit change in the cost/benefit equation could actually lead to an institutional increase in technology spending, due to greater demand. In other words, any savings derived from declining prices will likely be more in the form of cost avoidance rather than actual cost reduction. * When using the life-cycle approach to calculate annual hardware costs, care must be given to consider future computing trends. For example, in a commodity market where the price/performance ratio is dropping steadily, departments may find they can provide a superior computing environment over the long run by buying lower cost computers with shorter life cycles, rather than buying more expensive models and trying to make them last longer. * When calculating annual software costs, planners should consider the attendant support implications. Trying to support software that is either too new or too old can have a serious impact on support costs as well as the useability of the entire network. * During the planning process, departments should consider the functionality, type, range of computers, and number of software packages that will require support. Not all hardware and software are compatible, and supporting too many platforms, applications, or versions of the same program can have a significant impact on the amount of network support needed. In other words, the marginal cost of supporting an additional standard, software package, or platform can be prohibitively high. * Departments need to consider developing policies that clarify the department's position as to how much depth of support it desires versus the breadth of applications it intends to support. Budgeting for breadth is almost always more expensive than depth. * When planning for annual support costs, budgeting too little support can be more wasteful than budgeting too much. Systems that don't work reliably, or are too hard to learn or make work, often aren't used or waste large amounts of valuable time. * Planning for change is essential. Information technology changes regularly--adopting flexible and modular approaches to technology budgeting should pay in the long run (e.g., buying more smaller computers instead of fewer larger ones). * Although the capital cost of technology is dropping steadily, the support costs, organizational impact, and sociological changes inherent in implementing information technology remain significant. Delaying investments in information technology may only make these issues more difficult. Keeping current with technology and training is usually more efficient than a feast-or-famine approach. Reality check There are a host of issues beyond financial planning that departments need to consider when developing and implementing departmental technology plans. The following list is included only as a reality check to put the budget worksheet into perspective. * Technical Issues: Topology Standards Architecture Connectivity with administrative data processing * Governance: Access Security Service allocation Faculty vs. staff vs. students * Academic: Collaborative support Multimedia Classrooms and labs Integrating technology into the classroom Internet connectivity * Financial: Total budget needs Source of funds Life cycles Student fees Fee-based services Skeptics of budgeting and planning for information technology may still view any technology planning as an oxymoron. In some ways they may be right. Prediction will never be a perfect science, and predicting changes and future directions for information technology will always be particularly difficult. Moreover, it may be that many technology planners will continue to underestimate the cost of information systems, due to socially ingrained biases and an under-appreciation of the accelerating rate of change inherent in information technologies. Three forces at work that will continue to make planning and budgeting difficult are: (1) "expectation inflation," where both planners and users underestimate the desirability of future information systems that will deliver new and improved levels of performance; (2) "life-cycle optimism," where planners are coerced by their own false optimism as well as financial and management pressures to adopt an overly optimistic estimation of the true life cycle of technology investments; and (3) "investment creep," where financial pressures on institutions will increasingly force them to invest more in technology as it becomes a more cost-effective investment. Still, life-cycle budgeting offers much promise in helping all concerned recognize, address, and resolve these issues. Conclusion When reviewing the strengths and weaknesses of life-cycle budgeting, it is obvious that although this method has many strengths--improved coordination, identification of costs, clearer expectations, and improved understanding among faculty, departmental committees, and administrations--it also has weaknesses. The most prevalent weakness is that the method ultimately relies on making appropriate and realistic assumptions about quantities, prices, life cycles, and levels of acceptable performance and support. In an era of rapid change in technology and usage, these answers are not easily known. In this sense, life-cycle budgeting potentially shares a common characteristic with the systems it attempts to budget--garbage in, garbage out. Nevertheless, the fact that life-cycle budgeting makes these assumptions so clearly explicit gives it enormous potential. Over time, planners and institutions that adopt this methodology and understand the supporting assumptions should develop new levels of expertise that will serve them and their institutions very well. ************************************************************************ Sources of Help for Departments in Estimating Life Cycles Life-cycle budgeting is dependent on making good assumptions. Consulting and collaborating with peers and other professionals is an essential part of the assumption-making process. More sources of information are available today than ever before. The number of organizations available to help plan and implement technology has grown steadily over the last decade. Leveraging these organizations and individuals should be part of the planning and budgeting process. The following list represents only a small sample of the organizations available to assist in this process. Central Campus Computing Organizations Central computing organizations usually have offices that can offer consultancy services to departments. These organizations are usually the easiest places to obtain assistance. Other Campus Departments Consulting with other departments and peers is an excellent source of information. Examining their experiences and successes can be an excellent basis for a "reality check." Peer Institutions While other departments can be an excellent reference point for departmental plans, high-level campus administrators often look to peer institutions as a reference point. Including comparisons to peer institutions can be helpful when seeking new funding. Professional Organizations Many national organizations are a good source of information. Two of the best known technology organizations, CAUSE and EDUCOM, offer a variety of useful technology publications, and CAUSE offers a database service that enables peer comparisons. Professional organizations like the National Association of College and University Business Officers (NACUBO) and the Society for College and University Planning (SCUP) also address technology planning and funding issues. Industry Research Groups Industry research groups can be helpful in developing requests for proposals and gathering information on industry standards and trends (e.g., The Gartner Group, IDC, Dataquest). They offer one potential advantage over technology vendors in that they are supposedly without bias. Trade Press Many information technology trade publications regularly survey the industry for standards and print results from other studies. These publications can provide a wealth of timely information and references for further research. For these reasons, many information technology professionals regularly monitor such publications as ComputerWorld, Information Week, Networking, and CIO Magazine, to name just a few. Industry Vendors Technology vendors, like users, estimate product life cycles. In fact, they spend considerably more time planning and estimating them than do users. While they face uncertainties similar to those faced by users, vendors can be very helpful consultants regarding life-cycle estimates. The author has received valuable and insightful information from Apple, IBM, Zenith, and other vendors regarding technology life cycles. ************************************************************************ Bibliography: Benjamin, Robert I., and Jon Blunt. "Critical IT Issues: The Next Ten Years." Sloan Management Review, Summer 1992, pp. 7-19. Hawkins, Brian L. Organizing and Managing Information Resources on Campus. EDUCOM Strategies Series. McKinney, Texas: Academic Computing Publications, Inc., 1989. Norton, John A., and Frank M. Bass. "Evolution of Technological Generations: The Law of Capture." Sloan Management Review, Winter 1992, pp. 66-77. Ringle, Martin D. Computing Strategies in Liberal Arts Colleges. EDUCOM Strategies Series. Reading, Mass.: Addison-Wesley Publishing Company, Inc., 1992. Shank, John K., and Vijay Govindarajan. "Strategic Cost Analysis of Technological Investment." Sloan Management Review, Fall 1992, pp. 39- 51. ************************************************************************ John Oberlin, formerly Project Director of the Institute for Academic Technology, is the Director of Finance and Planning for the Office of Information Technology at the University of North Carolina-Chapel Hill. As such, he is responsible for implementing a broad-based financial planning effort for information technology, and for developing an integrated approach to the information technology planning process across units and academic departments. He holds a BA in quantitative economics from the University of California, San Diego, and an MBA from UNC-Chapel Hill. ************************************************************************ Departmental Budgeting for Information Technology: A Life-cycle Approach