复旦大学：《电子治理》阅读材料（英文）：Smart IT决策分析方法 Making Smart IT Choices.pdf_P11-P15

When new technology is brought into the manual record keeping, speed up transaction picture, a cumbersome and inefficient process increase revenue, prevent errors, or support may be automated"as is. The results are good and timely decisions. All of these kinds systems that do not serve business needs, of improvements should be represented by systems that are too expensive for the small baseline measures and target levels of productivity gains they provide, and systems performance. Without them, the technology that are not flexible enough to meet changing tends to take on its own independent course, demands. This leads to poor performance in and fails to have its intended impact on how individual organizations and even worse real people use information to accomplish real roblems when the system is expected to connect multiple programs or agencies, as is often the case in government Failure to understand the strengths and limitations of new technology In order to make system design work more Information technology is constantly changing anageable, we often ignore the ways in which and improving No one is able to keep up with one system affects related work processes. For the details of all new developments or to example, an accounting system needs to factor understand comprehensively how each new in the ways in which accounting is related to technical tool works. Add to this the fact that budgeting, revenue collection, and financial most new technologies must work in tandem anagement. The accounting function does with others, or must be incorporated into not stand alone, and a system that supports existing older systems, and the possibility accounting cannot behave as if it does for trouble mounts rapidly. Since most organizations are not in the IT evaluation Often organizations are not willing to invest business, they may rely on word of mouth the time and money necessary to do compre- vendor claims, and trade publications for the hensive process mapping and analysis. Many bulk of their knowledge about which tools organizations see this as an unnecessary may be right for which jobs. There is little effort; in place of it they share procedures opportunity to learn first-hand before making manuals(typically outdated and not reflective expensive, irrevocable decisions f what is going on)with system designers Unfortunately, this sets up a"pay me now Projects that are too specialized or or pay me later" proposition Ineffective or ambitious to manage successfully incorrect processes become embedded in the For years, oversight organizations like the US new system causing need for costly revisions General Accounting Office(GAO)have warned or manual "work-arounds"that defeat the against information technology initiatives that purpose of applying technology in the first characterize as "grand designs. These place are the projects whose scope is so large, time horizon so long, or design so unique that they Lack of measurable alignment between will almost certainly falter or fail. Such projects organizational goals and project objectives invite delays, unexpected complications, gaps Another risk factor involves the alignment(or in funding, management nightmares, and other lack of alignment) between organizational and problems. Instead, most experts recommend project objectives. The goal of IT adoption that systems be designed and deployed in should be to enhance or improve an modules and be built on standard technologies organization s ability to carry out its main using well-tested methodologies. This risk mission or business objectives. For instance, is relative and applies to large and small it should improve customer service, reduce organizations. a town or county system with a CENTER FOR TECHNOLOGY IN GOVERNMENT: MAKING SMART IT CHOICES

CENTER FOR TECHNOLOGY IN GOVERNMENT: MAKING SMART IT CHOICES 9 manual record keeping, speed up transactions, increase revenue, prevent errors, or support good and timely decisions. All of these kinds of improvements should be represented by baseline measures and target levels of performance. Without them, the technology tends to take on its own independent course, and fails to have its intended impact on how real people use information to accomplish real work. Failure to understand the strengths and limitations of new technology Information technology is constantly changing and improving. No one is able to keep up with the details of all new developments or to understand comprehensively how each new technical tool works. Add to this the fact that most new technologies must work in tandem with others, or must be incorporated into existing older systems, and the possibility for trouble mounts rapidly. Since most organizations are not in the IT evaluation business, they may rely on word of mouth, vendor claims, and trade publications for the bulk of their knowledge about which tools may be right for which jobs. There is little opportunity to learn first-hand before making expensive, irrevocable decisions. Projects that are too specialized or ambitious to manage successfully For years, oversight organizations like the US General Accounting Office (GAO) have warned against information technology initiatives that some characterize as “grand designs.” These are the projects whose scope is so large, time horizon so long, or design so unique that they will almost certainly falter or fail. Such projects invite delays, unexpected complications, gaps in funding, management nightmares, and other problems. Instead, most experts recommend that systems be designed and deployed in modules and be built on standard technologies using well-tested methodologies. This risk is relative and applies to large and small organizations. A town or county system with a When new technology is brought into the picture, a cumbersome and inefficient process may be automated “as is.” The results are systems that do not serve business needs, systems that are too expensive for the small productivity gains they provide, and systems that are not flexible enough to meet changing demands. This leads to poor performance in individual organizations and even worse problems when the system is expected to connect multiple programs or agencies, as is often the case in government. In order to make system design work more manageable, we often ignore the ways in which one system affects related work processes. For example, an accounting system needs to factor in the ways in which accounting is related to budgeting, revenue collection, and financial management. The accounting function does not stand alone, and a system that supports accounting cannot behave as if it does. Often organizations are not willing to invest the time and money necessary to do comprehensive process mapping and analysis. Many organizations see this as an unnecessary effort; in place of it they share procedures manuals (typically outdated and not reflective of what is going on) with system designers. Unfortunately, this sets up a “pay me now or pay me later” proposition. Ineffective or incorrect processes become embedded in the new system causing need for costly revisions or manual “work-arounds” that defeat the purpose of applying technology in the first place. Lack of measurable alignment between organizational goals and project objectives Another risk factor involves the alignment (or lack of alignment) between organizational and project objectives. The goal of IT adoption should be to enhance or improve an organization’s ability to carry out its main mission or business objectives. For instance, it should improve customer service, reduce

Limited authority to make decisions procurement are goals of integrity and fairness By design, governmental authority is divided the processes are often a source of problems cross multiple decision makers. Executive and delays. Agencies write managers do not have a clear line of Proposals(RFPs)using the information they authority over operations. Their decisions are have been able to gain from limited research circumscribed by law, limited appropriations, Vendors spend large sums of money trying to civil service rules, and a variety of legally develop the winning response Time consuming mandated procedures or court decisions arms-length reviews and negotiations ensue These restrictions do not blend well with the Losers often take advantage of bid protest complexities of managing an expensive and procedures that can further delay contract complex IT project in a rapidly changing awards for months-or even years. Due to technical environment Worse. as can be seen insufficient project support or understanding in the California DMV experience, when IT by top management, IT procurement requests projects fail the common legislative response may receive low priority. The resulting delays is to place more restrictions and more controls can mean time and cost overruns which in over the IT management process turn yield negative publicity, decreased support from top management, and negative Multiple stakeholders and competing goals perceptions of the overall value of the effort Government programs are characterized by a Commodity-based procurement, on the other multiplicity of stakeholders who often have hand, is easy for agencies to use, but assumes competing goals. Customers, constituents, that they have all the information and expertise taxpayers, service providers, elected officials, they need to design and assemble a high professional staff, and others all have some performance system out of a catalog of parts. stake in most programs. Some want more or different services. others want lower taxes or Little capability to design or operat fewer rules. Understanding how different integrated or government-wide programs choices may affect each stakeholder group Critics(including many public officials helps to identify likely conflicts and prevent themselves)complain that different unexpected problems government agencies operate more or less without regard to the fact that they often serve One year budgets the same people. The difficult fact of life is that Uncertainty about the size and availability government is organized mostly into separate of future resources weakens the ability of programs that receive specific authorization overnment agencies to successfully adopt and funding from Congress or a state lew IT innovations. Most government legislature. Accountability rules and tradition budgets are handled on an annual cycle demand that these programs be operated While many agencies have developed planning separately, and one result is the famous mechanisms to cover a three to five year stovepipe" systems of government. eriod, annual appropriations(influence Because these isolated programs emerge heavily by changing government-wide from deliberate design, they are very difficult As a result, funds promised for a project/ a priorities)tend to negate long-term plannit to coordinate, much less integrate New technologies, especially the Internet, are the first year may not be continued during the making it feasible to present a unified face second or sub to the public, but the changes that need to take place behind the scenes in policies, Highly regulated procurement accountability mechanisms, processes, data Most decisions to adopt emerging technologies sharing, and records management are only are made through the traditional competitive beginning to be understood bidding process. While the goals of competitive CENTER FOR TECHNOLOGY IN GOVERNMENT: MAKING SMART IT CHOICES

CENTER FOR TECHNOLOGY IN GOVERNMENT: MAKING SMART IT CHOICES 11 Limited authority to make decisions By design, governmental authority is divided across multiple decision makers. Executive managers do not have a clear line of authority over operations. Their decisions are circumscribed by law, limited appropriations, civil service rules, and a variety of legally mandated procedures or court decisions. These restrictions do not blend well with the complexities of managing an expensive and complex IT project in a rapidly changing technical environment. Worse, as can be seen in the California DMV experience, when IT projects fail the common legislative response is to place more restrictions and more controls over the IT management process. Multiple stakeholders and competing goals Government programs are characterized by a multiplicity of stakeholders who often have competing goals. Customers, constituents, taxpayers, service providers, elected officials, professional staff, and others all have some stake in most programs. Some want more or different services, others want lower taxes or fewer rules. Understanding how different choices may affect each stakeholder group helps to identify likely conflicts and prevent unexpected problems. One year budgets Uncertainty about the size and availability of future resources weakens the ability of government agencies to successfully adopt new IT innovations. Most government budgets are handled on an annual cycle. While many agencies have developed planning mechanisms to cover a three to five year period, annual appropriations (influenced heavily by changing government-wide priorities) tend to negate long-term planning. As a result, funds promised for a project in the first year may not be continued during the second or subsequent years. Highly regulated procurement Most decisions to adopt emerging technologies are made through the traditional competitive bidding process. While the goals of competitive procurement are goals of integrity and fairness, the processes are often a source of problems and delays. Agencies write Requests for Proposals (RFPs) using the information they have been able to gain from limited research. Vendors spend large sums of money trying to develop the winning response. Time consuming, arms-length reviews and negotiations ensue. Losers often take advantage of bid protest procedures that can further delay contract awards for months—or even years. Due to insufficient project support or understanding by top management, IT procurement requests may receive low priority. The resulting delays can mean time and cost overruns which in turn yield negative publicity, decreased support from top management, and negative perceptions of the overall value of the effort. Commodity-based procurement, on the other hand, is easy for agencies to use, but assumes that they have all the information and expertise they need to design and assemble a highperformance system out of a catalog of parts. Little capability to design or operate integrated or government-wide programs Critics (including many public officials themselves) complain that different government agencies operate more or less without regard to the fact that they often serve the same people. The difficult fact of life is that government is organized mostly into separate programs that receive specific authorization and funding from Congress or a state legislature. Accountability rules and traditions demand that these programs be operated separately, and one result is the famous “stovepipe” systems of government. Because these isolated programs emerge from deliberate design, they are very difficult to coordinate, much less integrate. New technologies, especially the Internet, are making it feasible to present a unified face to the public, but the changes that need to take place behind the scenes in policies, accountability mechanisms, processes, data sharing, and records management are only beginning to be understood

Extreme risk aversion save using IT, or even if you will save Governments business is public business. money at all. This book is about how to think This means that most new ideas have to be about an it investment It offers a set of implemented in full public view. An innovation analytical techniques to understand the issues gone wrong" risks not only dollars, but also and opportunities and to build a business case the credibility of an agency and its leadership for investing in a particular path. It shows how legislators, executive officials, and the to use that business case to get the support s not surprising then, that government you need to move forward with your project. to rely on the tried and true. We think of this phase of the IT investment process as the one that comes "before the How this guide can help beginning. " It is necessary before the first design meeting, before an RFP is written, Government managers need to analyze and before the budget is developed. We believe this evaluate IT choices because these choices kind of up front analysis is essential to doing are among the most complex and expensive all of those things well because it uncovers decisions they are expected to make. Whether both risks and resources that lead to smarter you are a local official considering a new IT decisions ystem to support building permits or an administrator at a large federal agency In short, three kinds of analysis mitigate the considering a new network infrastructure, risks of these investments you are faced with a complex and relatively a thoroughly understand both the problem to expensive decision-making process. The be solved and its context consequences of your IT decisions often have a significant and direct impact on the public a identify and test possible solutions to the For example, the safety of the flying public problem rests on the ability of the Federal Aviation Administration to implement systems that a evaluate the results of those tests against control air traffic. A state child welfare agency clear service and performance goals gathers and responds to information that protects the health and well being of children The following chapters present a well-tested in that state. A local government emergency methodology that can help you to understand response application ensures that emergency and carry out these three critical tasks vehicles are routed to incidents in the fastest and safest way possible. Systems that support References law enforcement make a big difference in the Hammer, Michael. (1990 )"Reengineering ability of federal, state, and local criminal Work: Don' t Automate. obliterate " Harvard justice agencies to provide public safety. These Business Review, July-August systems cost thousands, millions, even billions of dollars. They are important because the Isenberg, Phillip.(1994 )"Point of view When BIG IT Projects Falter, "Government goals they serve are crucial to our quality of life. The risks of system failure are linked to Technology, July, 7(7) the risks of service failure for hundreds Miller, Brian. (1994 )"DMV Project Hits thousands, or even millions of people Dead End, Government Technology, July, 7 7)pp. 1, 53-54.The Standish Group. (1995) This guide cant tell you what technology to The CHAOS Report On line at buy, or even what problem is most important to www.standishgroup.com/sample_research/ solve It cant tell you how much money you will chaos_1994_1.php 12 CENTER FOR TECHNOLOGY IN GOVERNMENT: MAKING SMART IT CHOICES

CENTER FOR TECHNOLOGY IN GOVERNMENT: MAKING SMART IT CHOICES 12 save using IT, or even if you will save any money at all. This book is about how to think about an IT investment. It offers a set of analytical techniques to understand the issues and opportunities and to build a business case for investing in a particular path. It shows how to use that business case to get the support you need to move forward with your project. We think of this phase of the IT investment process as the one that comes “before the beginning.” It is necessary before the first design meeting, before an RFP is written, before the budget is developed. We believe this kind of up front analysis is essential to doing all of those things well because it uncovers both risks and resources that lead to smarter IT decisions. In short, three kinds of analysis mitigate the risks of these investments: ■ thoroughly understand both the problem to be solved and its context ■ identify and test possible solutions to the problem ■ evaluate the results of those tests against clear service and performance goals The following chapters present a well-tested methodology that can help you to understand and carry out these three critical tasks. References Hammer, Michael. (1990.) “Reengineering Work: Don’t Automate, Obliterate,” Harvard Business Review, July-August. Isenberg, Phillip. (1994.) “Point of View: When BIG IT Projects Falter,” Government Technology, July, 7(7). Miller, Brian. (1994.) “DMV Project Hits Dead End,” Government Technology, July, 7 (7) pp. 1, 53-54.The Standish Group. (1995). The CHAOS Report. On line at www.standishgroup.com/sample_research/ chaos_1994_1.php Extreme risk aversion Government’s business is public business. This means that most new ideas have to be implemented in full public view. An innovation “gone wrong” risks not only dollars, but also the credibility of an agency and its leadership with legislators, executive officials, and the public. It’s not surprising, then, that government tends to rely on the “tried and true.” How this guide can help Government managers need to analyze and evaluate IT choices because these choices are among the most complex and expensive decisions they are expected to make. Whether you are a local official considering a new system to support building permits or an administrator at a large federal agency considering a new network infrastructure, you are faced with a complex and relatively expensive decision-making process. The consequences of your IT decisions often have a significant and direct impact on the public. For example, the safety of the flying public rests on the ability of the Federal Aviation Administration to implement systems that control air traffic. A state child welfare agency gathers and responds to information that protects the health and well being of children in that state. A local government emergency response application ensures that emergency vehicles are routed to incidents in the fastest and safest way possible. Systems that support law enforcement make a big difference in the ability of federal, state, and local criminal justice agencies to provide public safety. These systems cost thousands, millions, even billions of dollars. They are important because the goals they serve are crucial to our quality of life. The risks of system failure are linked to the risks of service failure for hundreds, thousands, or even millions of people. This guide can’t tell you what technology to buy, or even what problem is most important to solve. It can’t tell you how much money you will