System Error

Imagine there was a tax system in which you could:

  • File a single return which would provide data to both the Internal Revenue Service and your state.
  • Choose your own filing method: paper, touch-tone phone, or modem directly from your personal computer to the IRS.
  • Resolve issues or problems with a single phone call to an IRS customer service representative.
  • Get answers to tax questions through an extensive computerized data bank available 24 hours a day, 7 days a week.
  • Get a refund in two to three weeks rather than the current four to six weeks.
If the IRS Tax Systems Modernization project is completed, not only will the taxpayers of the future enjoy these benefits, but the agency will be able to:
  • Capture 100 percent of tax return data using automation instead of the 60 percent currently captured through automation.
  • Reduce error rates in data captured to less than 1 percent.
  • Use additional data to target compliance efforts, allowing a greater proportion of taxes owed to be collected.
The IRS estimates its modernized tax collection system could provide the government $40 billion per year in additional revenue without increasing tax rates. Millions of hours of taxpayer effort would be saved by reducing erroneous notices and providing quicker, more effective service.

At $7 billion, the new system's cost would be less than one-quarter of the estimated additional revenue that would be collected each year. It sounds like a good investment, but seven years and approximately $3 billion into the Tax Systems Modernization (TSM) program, the project is in jeopardy. The Treasury, Postal Service and General Government Appropriations bill approved by the House in July would have provided $425 million for TSM, half the amount requested by the President.

The House made apparent its lack of confidence in the IRS' ability to manage the project, fencing the entire amount pending action by the IRS to contract out program development and delivery to the private sector. In a remarkable slap at the agency, the House bill's language required that the restructuring of the contractual arrangement with the private sector be handled by the Defense Department. Whether that provision remains in the final version of the bill is in doubt, however. The Senate Appropriations Committee version did not include similar language. The funding cuts do appear likely to stick, as the Senate Appropriations Committee approved only $402 million for the project. The committee did provide more funds than the House for information systems, however, some of which may relate to TSM. Funding cuts in fiscal 1996 and the near certainty of further cuts led IRS to announce in July that it plans to cut 5,000 jobs in fiscal 1997. Information services, which includes TSM, will bear the brunt of staff reductions.

How a program most taxpayers would regard as common sense got to this point is a story that many in government will regard as all too familiar. It is a story of internal politics, talented managers thwarted by the system and bureaucracy impeding change.

The Most Powerful Bureaucracy

With a professional core composed mostly of accountants and a mission conducive to mass-processing and assembly-line techniques, the IRS is one of the most conservative and bureaucratic organizations in Washington. Its reputation among citizens is formidable. Author David Burnham calls the IRS "the single most powerful bureaucracy in the world" and "the single most powerful instrument of social control in the United States," in his 1989 book about the agency, A Law Unto Itself.

Within the federal government, the IRS is generally regarded as well-run. Financial World magazine, in a 1994 evaluation of management at 10 large federal agencies, gave the IRS the highest grade-a B+. "The IRS is advanced in strategic planning and also leads in financial reporting," the magazine said.

Ironically, the Tax Systems Modernization program, which Financial World cited as evidence of the IRS' strategic planning, has put the agency's reputation in serious jeopardy. Before the congressional efforts to cut the program this summer, three important oversight bodies strongly criticized management of the program. In a June report, the General Accounting Office said TSM is "at serious risk due to pervasive management and technical weaknesses." A panel convened by the National Research Council (NRC) of the National Academy of Sciences used similar language, saying, "TSM is in danger of being ultimately ineffective because in the IRS there is a lack of technical expertise and experience in managing large-scale system development." Leonard Baptiste, senior assistant director of accounting and information management at GAO, says TSM is over budget, behind schedule, and below expectations.

Dealing With the Rust Factor

By any measure, Tax Systems Modernization is a massive undertaking. A recent report from the Treasury Department to the House and Senate Appropriations committees lists 27 separate TSM projects costing anywhere from $30 million to $670 million each.

TSM began in the late 1980s as a project to update the agency's antiquated computers, some of which were as much as 20 years old. The "rust factor" was so severe, according to officials, that the computers were in danger of literally falling apart. "The driving point for TSM was the tax processing system," says Michael Murphy, IRS deputy commissioner at the time, and now director of the Tax Executives Institute. "The pressure was the processing people coming to me and the commissioner and saying, 'This thing is being held together by a thread; we are not sure how much longer we can keep the system together to process returns.' "

A proposal to simply upgrade the computers and automate existing processes was rejected in part because of GAO and NRC recommendations. They argued that automating existing processes would capture only 10 percent of the project's potential benefits. Both groups recommended the IRS consider completely restructuring the way it does business.

Consistent with those recommendations, IRS leaders decided to couple the massive systems modernization project with a complete overhaul of the agency's business processes. The key to the new approach was the move from a paper-based system for processing tax returns to electronic methods. Modern information technology, including scanners, voice response units and laptop computers were to become part of the agency's way of doing business.

Significant operational changes were to accompany the technological advances. For example, highly specialized jobs such as revenue agent, taxpayer representative and revenue officer were to be combined into a single generalist position of customer service representative. The intent was to enable a taxpayer to resolve a problem with a single phone call to a customer service representative rather than calls to multiple individuals and offices.

Tensions Over Modernization

Joining modernization with a fundamental change in processes created a power struggle between IRS's information systems division and the key operations units-examination, collections and return processing. The information systems division initiated TSM to modernize the agency's computers, but operations units needed to take the lead on changing business procedures. "To make the system work right, operations has to control it," says David Blattner, former chief operations officer and a key player in the early days of the project. "It is being designed to produce operations results, not system modernization results."

Operations officials were reluctant to get involved because they lacked confidence in the information systems division. "Systems said, 'We will deliver.' They never did, and when they did it wasn't for priority needs that operations people had identified," says Phil Brand, a former IRS chief compliance officer.

As an example, Brand cites the consolidation of 70 telephone sites into 26. "The announcement was made and the sites were closed but there was no place to send the work. The technology was not there to integrate the call sites. They tried to do too much beyond their capability."

Henry Philcox, former IRS chief information officer and head of information systems, has a different perspective. "I had a lot of difficult issues to work through to gather requirements. The user had to define the needs. We had problems in pulling that together," he says. "I was a peer in that process. I needed resources from the operations units to make the next phase work. They would say, 'I don't have the staff and can't provide the people.' "

In an attempt to get the operations side more involved, the Office of Modernization Executive (ME) was created, and former Austin Service Center Director Larry Westfall was named to head it. But creation of the ME's office failed to solve the problem. The CIO, not the ME's office, continued to control TSM funds. "Nobody was in charge. Larry was a matrix manager without line authority. He tried to cajole and persuade," says a former modernization office staffer.

Eventually, at the insistence of Congress, the Modernization Executive's office was given control of the budget. In the fiscal 1995 Treasury appropriation, the House Appropriations Committee required the IRS to give the ME's office authority and responsibility for the project, including control over the budget. The change enhanced Westfall's control over the project but did not resolve all the internal tensions.

Organizational Obstacles

Part of the problem was cultural. The IRS's executive committee, consisting of the heads of the six major operating units, the commissioner and the deputy commissioner, has traditionally operated by consensus. According to Westfall, consensus decision-making was not well-suited to the modernization project. "When you are running a change agenda, consensus takes too long. There are too many competing agendas," he says. "My personal view was it was necessary for strong authority to be effected out of a central command structure." Brand agrees. "There is nothing wrong with collegiality but there is also nothing wrong with learning how to salute," he says. Philcox concurs, saying, "What we needed was strong leadership. Everyone has to come out of the room and be focused."

Deputy IRS Commissioner Michael Dolan is the top career employee in the agency, and would presumably be the source of that leadership. Dolan sees the problem not as one of operating style but as the fundamental difficulty of balancing "the huge demand by the business side" for resources against the need to invest in technology. Dolan says unpredictable budgets for the project have caused an "endless recycling of choice-making" between these competing demands.

One option would have been "to take it in the ear on short-term operational issues" as one official described it, in order to focus energies and resources on TSM. However, history and what officials describe as "the tremendous press of daily business" foreclosed that option. During the infamous 1985 filing season, refunds were delayed because of problems with new computers and the IRS came under severe criticism from Congress for its performance. The highest priority has been given to preventing that situation from reoccurring.

Overcoming functional divisions in the agency has been difficult. The key functional units-returns processing, examination and collection-are at the heart of the agency's ability to deliver services. The strength of those units, however, lies in part in their insularity, which has been an obstacle as Westfall and others tried to make operational changes. "The functional organizations have spent years building system support structures that work and provide a strong program base," Westfall says. "When you run a change agenda on top of that, it blurs functional lines and modifies processes in a cross-cutting way." The functional units resisted strongly, Westfall says.

IRS Insularity

Cultural and organizational issues are seen by some as the heart of IRS' technical problems. For example, both GAO and the NRC panel criticized the IRS information systems division for being incapable of developing software to meet industry standards or adequately overseeing software development by outside contractors. "Although Tax Systems Modernization is much needed, the IRS does not currently have a development organization capable of meeting the difficult challenge," the NRC panel said in its final report on TSM.

Much of the problem derives from IRS' insularity and the insistence on doing everything in-house, says Robert Clagett, retired professor of business administration at the University of Rhode Island and chair of the NRC panel. "In the early 1990s, we said, 'You've got to get people with experience in software design, systems design and project management,' " Clagett says. "They finally got a few people who were systems designers. They treated them as outsiders. They never enmeshed them in the IRS and in design."

The systems designers had been brought in to assist in the creation of a systems architecture-an overall set of common standards for data and technology and a map of work processes and information flows. The lack of such a blueprint for the project has been repeatedly criticized by the NRC, GAO and Congress. "IRS has not completed its integrated systems architecture and has not committed to a completion date. . . TSM systems continue to be developed without the detailed architectures and discipline needed to ensure success," the June GAO report says. "If you don't have data architecture to tell the systems designers how data is captured and stored, you are leaving it to the designers to come up with interim solutions," GAO's Baptiste says. When incompatible components are integrated, the system could crash, he adds.

IRS Deputy Commissioner Dolan says it's "disingenous" to suggest that the IRS has failed to produce a systems architecture. "We have an architecture," he says. "It is not as complete as needed, but the inference that we have gone off and put out projects that can't connect is bull. The notion that we drug our knuckles, that there is no architecture, is not a total or accurate story."

Dolan also takes issue with the GAO and NRC argument that an outsider should have been brought in as modernization executive. When Westfall retired last fall, the former head of returns processing, Judy Van Alfen, was appointed to the position. "She is a good manager; she has been in several parts of IRS but she has no project experience whatsoever," says NRC's Clagett. "There is nothing in her background to equip her to do the job."

TSM has such important implications for how the IRS operates that "we really needed somebody who understands the business," Dolan responds. He points out that the new CIO, Arthur Gross, was brought in from the outside, having headed a similar modernization effort in New York state.

Overcoming Obstacles

Many of the obstacles to TSM relate to the general nature of government: pay scales that preclude offering salaries sufficient to attract talented outsiders, procurement lags of 2 to 4 years in delivery of large systems, constant demands by oversight groups, and the need to plan a 10-year project around the annual congressional appropriations process.

Nevertheless, Clagett points out, other agencies have overcome some of these problems. The Defense Department has acquired significant expertise in developing large systems. "We do buy weapons systems that work, but DoD doesn't build them. They set rigid specifications for performance," Clagett says.

Dolan doesn't buy the DoD analogy. "You can't compare this to Defense procurement," he says. "It is not a turnkey system. You have to migrate the master files; you aren't doing something that is severable. That has added to the burden of delivery."

A Success Story

Dolan argues that "truckloads of business value" provided by TSM have been largely ignored by critics. One notable success is the Integrated Collection System, which has greatly increased productivity among IRS collection officers. Under the old system, revenue officers trying to track down delinquent accounts would work from paper files and would have to visit the Department of Motor Vehicles or the Credit Bureau to get information, says Mark Gray, formerly assistant director of negotiations at the National Treasury Employees Union, who helped implement the new system. With laptop computers, all necessary information is now at the revenue officers' fingertips, he says. "Previously, when the revenue officer would knock on the door and the taxpayer said, 'The check is in the mail,' there was no way to verify the information," says Gray. Now, revenue officers will have up-to-date information on deposits. The officer can punch up a taxpayer's name on a computer and see whether his check bounced. "We have seen 30 to 40 percent increases in revenue collection in the areas that are using it," Gray says

Former IRS Commissioner Fred Goldberg, who helped convince Congress to significantly increase funding for TSM in the early 1990s, says the key benefit of the program is saving taxpayers time and aggravation. Goldberg says the accuracy of the information being captured electronically has spared millions of Americans the irritation of erroneous notifications. "The psychic value of not getting an incorrect notice is huge," he says. "Take it to the bank that it has happened because of TSM."

Philcox emphasizes great progress in replacing antiquated computers. "There is no comparison between how the IRS looks and its computer capabilities now and when I came in 1986," he says. "We were within a hair of not being able to support the tax system. We have huge capabilities today."

Change Is Messy

The TSM story underscores the messiness of organizational change. It is replete with clashes of personality, culture, power and politics. GAO reports may be misleading when they suggest that change is a neat, linear process and that more careful planning is the solution to problems with TSM. As Westfall points out, "nobody has the answer in a neat little package."

That caveat may apply as well to the Treasury Department's report to Congress on TSM. Treasury recommends IRS contract out the system's integration and software development and give Treasury a more prominent oversight role. It will take at least two years to get a prime integration contractor on board, a critical period during which the project will be in limbo. Further, there is a question whether the IRS can adequately manage its contractors. According to GAO's June report, the Cyberfile project, a contractor-run effort to allow returns to be filed via modem from personal computers, "was not developed using disciplined management and technical practices" and "was not ready for planned testing during the 1995 filing season."

To enhance IRS technical capabilities, Treasury has worked with OMB to create the Presidential Technology Team (PTT) of technical experts from across government that the IRS will "borrow" to help with TSM. Ultimately, however, success may hinge on whether progress is made in changing the old cultural bugaboos. IRS will have to overcome its traditional hostility toward outsiders in order to take advantage of the technology team's assistance and functional units will have to compromise their traditional independence and defer to strong central leadership on a change agenda.

Stay up-to-date with federal news alerts and analysis — Sign up for GovExec's email newsletters.
JOIN THE DISCUSSION
Close [ x ] More from GovExec
X CLOSE Don't show again

Like us on Facebook