In the past turbulent decade, many companies have been forced to recognize that they weren't organized in the right way to do what they were doing. Their organization structure reflected history, not current needs. Reform wasn't easy--too many people had vested interests in preserving their particular part of the organization. As a result, most attempts at reorganization were reduced to shifting things among different boxes on organizational charts. Businesses found that the only way to break the mold was to reengineer--to forget how they were organized, decide what they needed to do, and design the best structure to do it. An obvious insight? Perhaps. But the best ideas are always the ones that seem obvious--after their discovery.
We are determined to move from an industrial age government to information age government, from a government pre-occupied with sustaining itself to a government clearly focused on serving the people.
Vice President Al Gore
May 24, 1993
We will reengineer the work of government agencies in two ways. First, we will expand the use of new technologies. With computers and telecommunications, we need not do things as we have in the past. We can design a customer-driven electronic government that operates in ways that, 10 years ago, the most visionary planner could not have imagined.
Second, we will speed up the adoption of new ways to improve federal operations. Most of this work will be done by the federal agencies themselves. An outside performance review could never learn enough about internal agency work processes to redesign them intelligently. But we can begin to redesign several broad government-wide processes: The way we design programs, develop regulations, and resolve disputes.
The history of the closing decade of this century is being written on computer. You wouldn't know it if you worked for many federal agencies, however. While private businesses have spent the past two decades either getting rich by developing new computer technologies or frantically trying to keep up with them, government is still doing things our parents--perhaps even our grandparents--would recognize.
Offshoots of the unexpected and fertile marriage between computers and telephones have changed just about everything we do--how we work, where we work, the design of the workplace, and the skills we need to continue working.
Organizations don't need as many people collecting information because computers can do much of it automatically. They don't need as many people processing that information because clever software programs can give managers what they need at the press of a button.
Factories don't need to stockpile large inventories because smart machines on the assembly lines order components from equally smart machines working for suppliers. Yet government agencies stand guard over warehouses of unused office furniture. Retailers ship the right size of clothing to customers as soon as they receive a telephone order and a credit card number. Yet we can't pay our taxes that way.
Computer companies give technical advice for our computers and software over the telephone 24 hours a day by fax, modem, or voice. Yet, the Social Security Administration can't do the same.
Failure to adapt to the information age threatens many aspects of government. Take the State Department, a globe-spanning organization dependent on fast and accurate communications. Its equipment is so old-fashioned that the Office of Management and Budget says "worldwide systems could suffer from significant downtime and even failure." See Note 60 According to OMB, its systems are so obsolete and incompatible that employees often have to re-enter data several times. These problems jeopardize our ability to meet our foreign policy objectives.
Or think about the way our government sends out checks. For 15 years, electronic funds transfers have been widely used. They cost only 6 cents per transfer, compared with 36 cents per check. Yet each year, Treasury's Financial Management Service still disburses some 100 million more checks than electronic funds transfers.
We still pay about one federal employee in six by check and reimburse about half of travel expenses by check. Only one-half of Social Security payments--which account for 60 percent of all federal payments--are made electronically, making SSA the world's largest issuer of checks. Only 48 percent of the Veterans Affairs Department's payments are made electronically. Fewer than one in five Supplemental Security Income payments and one in ten tax refunds are transferred electronically. See Note 61 We have only begun to think about combining electronic funds transfers for welfare, food stamps, subsidies for training programs, and many other government activities.
Private financial transactions have become a lot easier in the past decade: bank cash machines are open 24 hours a day, credit cards let us avoid carrying cash, and we can buy goods over the telephone. This saves many of us a lot of time and money. It could save the Government a lot of time and money, too. Consider the paper chase involved in running the welfare system. The Food Stamp Program, alone, involves billions of bits of paper that absorb thousands of administrative staff years. More than 3 billion food stamps will be printed this year and distributed to more than 10 million households. Each month, 210,000 authorized food retailers receive these coupons in exchange for food. These retailers carry stacks of coupons to 10,000 participating financial institutions, which then exchange them with Federal Reserve Banks for currency. The Federal Reserve Banks count the coupons--although they already have been counted more than a dozen times--and destroy them. The administrative cost of this system--shared equally by federal and state governments--is almost $400 million a year.
We will support Agriculture's commitment to the goal of issuing food stamps electronically by 1996. Electronic benefits transfer could eliminate the paper chase, improve services to customers, and reduce fraud. At the same time, it could be used to authorize Medicaid payments, distribute welfare payments, infant nutrition support, state general assistance, and housing assistance. It could eliminate billions of checks, coupons, and all the other paperwork, record keeping and eligibility forms that clutter the welfare system.
Why has business moved faster than government into the electronic marketplace? In the first place, government is a monopoly. Public organizations don't go out of business if they don't have the latest and smartest machines or the best approach to managing resources. In the second, employees who do want to modernize management have their hands tied with red tape--detailed budgets and cumbersome procurement procedures-- that deter investment. Finally, there is a natural inclination, familiar to private and public managers alike, to do things as they've always been done.
What can we do to help our federal bureaucracy catch up?
OMB has already begun the process. The electronic benefits transfer steering committee, which OMB oversees, will develop an implementation plan for electronic benefits transfer by March 1994.
The system is workable with today's technology. For cash programs such as federal retirement, social security, unemployment insurance, or AFDC, benefits would be electronically deposited directly into recipient bank accounts electronically. If people didn't have bank accounts, these could be created once the individual enrolled in a program. For "non-cash" programs such as food stamps, participants would have accounts through which they could make purchases at approved food stores--analogous to credit cards with credit limits. Stores would debit accounts as eligible items were purchased. The entire system could operate on or be compatible with the existing commercial infrastructure through which private funds are transferred electronically.
Agencies have begun experiments with electronic benefits transfers. Welfare checks, food stamps, and state-collected child support, for example, are distributed electronically in Maryland. There are test sites in Iowa, Minnesota, New Mexico, Ohio, Pennsylvania, Texas, and Wyoming. We know that a joint federal-state effort to transfer welfare benefits electronically works--and works well. The system is strongly supported by recipients, the state welfare agencies, food retailers, banks, and participating commercial networks. We also know that direct federal delivery of funds by electronics is cost-effective. We can't yet project with certainty what the savings might be, but preliminary estimates suggest $1 billion over five years once electronic benefits transfer of food stamps is fully implemented.
In the future, the concept of electronic government can go beyond transferring money and other benefits by issuing plastic, "smart" benefit cards. With a computer chip in the card, participants could receive public assistance benefits, enroll in training programs, receive veterans services, or pay for day care. The card would contain information about participants' financial positions and would separately track their benefit accounts--thus minimizing fraud. Electronic government will be fairer, more secure, more responsive to the customer, and more efficient than our present paper based systems.
Barriers still stand in the way. Agencies will have to work together to develop a comprehensive nationwide strategy for implementation; it will do no good for each agency to develop its own process. We will need to strengthen the partnership between state and federal governments in developing and operating the system. We will have to eliminate some regulations that would prevent this radical change in how government operates. And the National Institute of Standards and Technology will have to issue final standards and protocols for electronic signatures to facilitate electronic funds transfers and the electronic approval of budget and financial documents.
Opportunities abound for cutting operating costs by using telecommunications technologies. The National Performance Review has identified several projects that would improve government's productivity and reduce the burden of reporting on individuals and businesses.
The IRS is introducing an efficient computer system, automating tax returns, and creating a wholly new work environment for its 115,000 full-time personnel. The agency currently operates a computer system put together in the 1960s--not the tool our principal revenue collector should be using. To make the new system work, the agency will need to figure out how to train its staff to operate in a reengineered agency. We will support the agency's investments in new hardware and training, as discussed in Chapter 3.
The IRS will also manage the creation of an integrated electronic system for financial filing, reporting, and tax payment by 1996. The system will serve federal, state, and local taxpayers. It will allow the electronic filing of tax returns by individuals and companies, the electronic reporting of wages and withholding information, and other data required by all levels of government. In addition, the inter-agency Wage Reporting Simplification Project (WRSP) will be in place quickly--allowing businesses to file information once to serve many different purposes. The savings from fully implementing this program over the life of the system have been projected at $1.7 billion for government agencies and $13.5 billion for private employers. Individuals will be able to file federal and state income taxes simultaneously through an Electronic Data Interchange, with their privacy protected and fraud prevented through digital signature standards. Electronic filing alone will save the IRS and state agencies from having to mail out the equivalent of 75 boxcars of forms.
Working together, the Labor Department and IRS will develop an automated system all employers can use to file electronically the pension plan forms employers required by the Employee Retirement Income Security Act. See Note 64 At present, it costs the Internal Revenue Service more than $10 million a year to enter all these forms into its data base.
The Labor Department will develop computer programs to determine quickly the appropriate wages on federal service contracts. See Note 65 Currently, all federal agencies contracting for services--from cleaning services to building management--must apply to the department for a determination of appropriate wages. The process is supposed to ensure that federal contracts don't undermine local prevailing wages. The process takes an average of 57 days and, with a growing number of service contracts, more and more are subject to delays.
We will continue investing in the Social Security Administration's massive project to create a single nationwide disability processing system.66 This will require considerable investments in new telecommunications and computer systems as well as in staff retraining. It will also mean that the SSA will have to work cooperatively with state-run disability determination offices, set performance standards, and take over those that don't meet standards. Many of the system's worst processing bottlenecks are in the state offices that approve individual claims.
Money for Numbers
The National Technical Information Service runs a large and complex information collection and marketing operation. It is the nation's largest clearinghouse for scientific and technical information. Yet it covers the costs of its operations without receiving a penny in federal appropriations. Its customers pay -- and their numbers are growing every year.
NTIS's archives contain about 2 million documents (from research reports to patents), more than 2,000 data files on tape, diskette, or CD-ROM, and 3,000 software programs. This resource is growing at the rate of about 70,000 items each year. NTIS's press releases, on-line services, and CD-ROMs serve 70,000 customers, three-quarters of whom are from business and industry.
In 1991, NTIS collected $30.7 million in revenues -- 77 percent from its clearinghouse activities, the rest from other government agencies that reimburse NTIS for patent licensing services, and from billing other agencies for producing and distributing documents. NTIS is required by law to be self-sufficient.
Some of these investments will require Congressional appropriations. But some can be financed through the innovation funds, described above, and some will become possible to pay for as soon as rigid budget regulations are relaxed.
Federal agencies must treat the data they compile and process as potentially valuable resources. Congress alerted the bureaucracy to the value of information in 1991 by passing the American Technology Preeminence Act. The act required federal agencies to transfer to the National Technical Information Service copies of federally funded research. At NTIS, the information is organized and made available to research scientists in academia and in industry. NTIS has developed an aggressive marketing strategy and pricing policy that have greatly increased its revenues.
The Census Bureau has pioneered the use of computer technology such as CD-ROM technology to make federal data available. By 1992, the Bureau sold census data to 380,000 customers on tape or disc directly, and served another 1.1 million customers indirectly.
Unfortunately, some federal agencies lag behind private data retailers in the services they offer their customers. People buying Census data must order it through paper order forms or by telephone during business hours--only 9 hours a day, 5 days a week. If private software companies offer 24-hour a day technical support, so should the Census Bureau.
Other agencies will begin to exploit the potential of the information they collect. The Commerce Department, for example, will develop a manufacturing technology data bank that brings together information residing in the National Institute of Standards and Technology, the Defense Department, federal research laboratories, and other organizations. Commerce will also use its climate data as the basis for developing a National Environmental Data Index. Good data will be vital in solving the problems associated with global climate changes. The U.S. must be a leader in developing these information resources.
Dozens of agencies collect spatial data--for example, geophysical, environmental, land use, and transportation data. They spend $1 to 3 billion a year on these efforts. The administration will develop a National Spatial Data Infrastructure, (NSDI) to integrate all of these data sources into a single digital resource accessible to anyone with a personal computer. This resource will help land developers and conservationists, transportation planners and those concerned with mineral resources, and farmers and city water departments.
Because of the value of the data, it will be possible to attract private sector funding for its collection, processing, and distribution. The Federal Geographic Data Committee, which operates under the auspices of OMB, plans to raise enough non-federal funding to pay for at least 50 percent of the project's cost. It will set the standards for data collection and processing by all agencies to ensure that NSDI can be developed as economically as possible.
The Customs Service lets people pay duties on imported goods by credit card. Americans should have the same convenient way to pay taxes. It will save time and cut the IRS's collection costs. See Note 70 There is one hitch: Those who pay by credit card could avoid paying back taxes simply by filing for personal bankruptcy. This escape mechanism can't be employed today because back taxes are, under bankruptcy law, a "non-dischargeable" debt--that is, they are a debt that remains even after someone becomes insolvent. Therefore, the use of credit cards for tax payments should be delayed until Congress has amended the bankruptcy statute to prevent taxes paid by credit card from becoming a dischargeable debt. Our goal is to increase customer convenience, not to open up another loophole through which people can dodge paying delinquent taxes.
Reengineering to Use Cost-Cutting Tools
Our reinvented government will be able to cut further costs by using new ways to carry out traditional duties. To begin with we will have to get a lot smarter about how we design government programs. The President's Management Council will play a lead role in helping government learn from its past failures and successes to design better programs. In addition new approaches to regulation--such as negotiated rule making-- can reduce conflict and produce better results. Finally, alternative techniques for resolving disputes can avoid many of the costs of traditional litigation.
As taxpayers and customers we have been, time and time again, victims of the thoughtless expansion of government. When new programs were introduced or old ones retargeted, little thought was given to what economists blandly label "second order effects"--the unintended and unwanted consequences of actions. These unintended consequences are the collateral damage responsible for so much of the waste documented in this report. When we placed limits on crop deficiency payments, we didn't realize how easy it would be to establish eligible shell-corporations. When we added new procurement standards, we didn't anticipate the difficulties caused by centralized decision making. When we tried to target training programs on dislocated workers, we didn't anticipate the bureaucratic hassles involved in establishing eligibility.
But the fact that we did not anticipate consequences does not mean that we could not have done so. Many different programs have been tried--by federal agencies, by state and local agencies, and by governments overseas. We have built up what lawyers would call "case law": lots of useful precedents about what works and what doesn't. The trouble is that, unlike case law, these precedents aren't easy to find. Congressional staff or agency employees designing new programs have no systematic way to find out what has been tried before and how well it has worked. The result? Endless reinvention of third rate or failed programs.
In 1981, for example, the chairman of the House Banking Committee asked the Congressional Budget Office if it knew of any studies evaluating government loans as an effective policy tool. CBO did not. Yet the federal government had lent hundreds of billions of dollars--and it continues to do so today. The price we pay for this ignorance is a mountain of delinquent debt and a raft of discredited government initiatives. Too many policies and programs are built on equally feeble foundations.
In 1988, Congress recognized this dilemma and provided for the establishment of a National Commission on Executive Organization, patterned after the first Hoover Commission. Its charter would have included a requirement to "establish criteria for use by the President and Congress in evaluating proposals for government corporations and government-sponsored enterprises and subsequently overseeing their performance." See Note 71 The new commission could have been activated by directive. It was not.
To begin our attack on ignorance, the President should direct the President's Management Council to make program design a formal discipline throughout the federal government. The PMC will commission the preparation and publication of a program design handbook and establish pilot efforts within agencies to strengthen their ability to design programs. These pilot programs will help senior management design new programs, evaluate current programs, and create models for many different types of programs (research contracts, loan programs, tax preferences, and insurance programs to name just a few.)
Since many programs originate in Congress, the Legislative branch should also work to improve staff capacity. We urge the Offices of the Legislative Counsel, the Congressional Research Service, and the General Accounting Office to fill this role. As both the legislative and executive branches elevate the discipline of program design, we will get better programs and less contentious relations between the two branches of government.
But we need more than good programs. We need better rules and more efficient rulemaking. Federal agencies administer tens of thousands of laws, rules, and regulations--and the number is growing quickly. For better or worse, government's rulemaking, even more than its appropriations, shapes our lives.
Costs, for the most part, are offset by benefits. Our system of laws and rules is the foundation for our economic success. It defines and protects personal and property rights and provides the framework for the orderly conduct of social and business affairs.
But some aspects of rulemaking don't work well. As rules extend into increasingly complex areas of our environment, workplace safety, health, and social rights, their consequences--both deliberate and unintended--also grow. As this happens, we introduce more and more safeguards into the rulemaking process. The result is not always what we want. Hearings, reviews, revisions, more reviews, more hearings, and even more reviews are cumbersome, costly, and time consuming. For example, because the Department of Health and Human Services has been slow to issue regulations on such vital areas as the allocation of funds for the elderly and for children, states have had to introduce their own regulations without the benefit of federal guidance. Some of these state regulations have later been overturned after federal regulations were eventually issued, leaving states financially liable.
New rules and regulations can also generate costly litigation--a bonanza for lawyers. Agencies writing the rules to implement environmental laws, according to one expert, often find "too frequently that their proceedings become a battleground for interest groups and other affected parties--in effect little more than the first round of the expected litigation." See Note 72
There are better ways to make rules. A small group of federal agencies has pioneered a process called negotiated rulemaking. In 1990, Congress recognized and encouraged the process with passage of the Negotiated Rulemaking Act. We believe negotiated rulemaking--colloquially referred to as "reg neg"--is a process every rulemaking agency should use more frequently. See Note 73
The "reg neg" process brings together representatives of the agencies and affected groups before draft regulations are issued and before all sides have formally declared war. The group meets with a mediator or "facilitator." The negotiators reach consensus on the regulation by evaluating their own priorities and making trade-offs. The negotiating process allows informal give and take that can never happen in court or in a public hearing. If agreement is reached, the agency can publish the proposed rule, accompanied by a discussion of the issues raised during negotiations. Even if both sides are too far apart to reach consensus, agency staff learn a lot during the process that helps them improve the regulations. When the parties do reach consensus, regulations are issued faster and costly litigation is avoided.
When EPA applied reg neg techniques to the issue of emission standards for wood burning stoves, it was able to put standards into effect two years faster, and with much better factual input, than it could have without negotiations. Manufacturers of stoves, in turn, were able to begin retooling to meet standards without another two years of uncertainty.
Federal agencies also need better and cheaper ways to resolve disputes. Enforcing thousands of difficult and sometimes controversial rules--however carefully they are designed--leads to disagreements. State and local governments, businesses, and citizens challenge Washington's right to regulate certain issues, or they challenge the the enforcement of specific regulations.
Solving these disputes can be expensive. It involves high-priced lawyers, it clogs the courts, and it delays action. Each year, 24,000 litigation matters reach the 530 full-time attorneys and 220 support staffers employed by the Labor Department alone. It often takes years to resolve these disputes, postponing the implementation of important programs and preventing a lot of people from doing what they are paid to do.
In some cases, litigation is important: it interprets the law, sets important precedents, and serves as a deterrent to future wrongdoing. But in many cases, no one really wins-- and the taxpayer loses. It is often cheaper to resolve conflicts through new techniques known collectively as Alternative Dispute Resolution (ADR).
Alternative Dispute Resolution (ADR) includes mediation (a neutral third party helps the disputants negotiate), early neutral evaluation (a neutral, often expert, person evaluates the merits of both sides), factfinding (a neutral expert resolves disputes that arise over matters of fact, not interpretation), settlement judges (a mediator settles disputes coming before tribunals), mini-trials (a structured settlement process), and arbitration (an arbitrator issues a decision on the dispute).
Overcrowded courts are already encouraging private litigants to use ADR. Private contracts often specify the use of ADR to resolve disagreements among signatories. In 1990, Congress passed the Alternative Dispute Resolution Act, authorizing every federal agency to develop its own ADR policy. Some have, but some have dragged their feet.
Those that have used ADR have saved time and money and avoided generating ill will. The Labor Department started a pilot program last year for OSHA and Wage and Hour cases and found it much quicker and cheaper. The Federal Deposit Insurance Corporation saved more than $400,000 with a single, small pilot program. The Farmers' Home Administration has used ADR on foreclosure cases--not only saving money but actually avoiding foreclosure on several families. This type of innovation should spread faster and further across the federal government.