Irs To Shut Down Several Technology Modernization Projects

[MORTARDUDE]

IRS TO SHUT DOWN SEVERAL TECHNOLOGY MODERNIZATION PROJECTS


OCTOBER 21, 2003


Hamilton, Amy
Tax Analysts

The IRS is expected to announce before the end of 2003 that it will shut down virtually all long-term technology modernization projects except those of the highest priority and will reorganize the structure for managing the massive $ 8 billion effort.

The IRS Oversight Board -- whose members are disappointed with cost overruns and delays in the modernization effort, according to the staff's director -- will help provide long-term guidance as the IRS determines how it will go from here.

In his first major interview since taking office, IRS Commissioner Mark W. Everson told The New York Times, "The effort to update the master file system has not been successful at all. That's something we have failed at." (David Cay Johnston, New York Times, Oct. 16, 2003.)

The IRS needs to replace its four-decades-old master file of almost 200 million taxpayer records and convert them into an information technology relational database. Only by replacing this relic from the Kennedy administration will the IRS be in a position to start functioning like any other financial institution with access to constantly updated account information. However, this summer the IRS again announced that implementation of the first phase of this new taxpayer account data engine will be delayed -- this time until at least the 2004 filing season.

The day after the Times interview with Everson appeared, John M. Dalrymple, the IRS's deputy commissioner for operations and support, met with IRS technology modernization project managers and indicated to them that bad news is coming. Everson created the deputy commissioner position held by Dalrymple to "own the modernization program" as supervisor of the IRS's chief financial officer, chief information officer, and chief human capital officer and the IRS's information technology and physical security operations.

"It's going to be a painful, painful process," Dalrymple said in October 17 comments to the Association of Former IRS Executives. He said this is because every IRS manager believes his or her project is vital to the agency's long-term technology modernization -- and the IRS's "appetite for improvement projects is insatiable."

"But when you stretch yourself out over all these projects you never finish the projects most important to you," Dalrymple said. "We're stretched too thin. We will have to shut down smaller projects. We've not done that before."

He added that the IRS's new organizational structure will help in this regard. Before, it was easy for a manager whose pet project had been cut to "come in the back door" by going to another IRS district and getting funding for it through another hierarchy. No one would notice until it was too late, Dalrymple said. The agency's new structure makes maneuvering like that pop out in the budget numbers, he said, adding that he supervises all technology-related IRS offices.

'Do Not Start Over'

In the mid-1990s one Republican House leader called that era's failed IRS technology modernization effort a "$ 4 billion fiasco," which led to accusations from lawmakers that the IRS was throwing money down a "rat hole." As a result, Congress required the IRS to partner with computer experts from the private sector in its current effort to integrate modernized technology systems across the agency.

In December 1998 the IRS awarded the prime contract to a consortium headed by Computer Sciences Corp. (CSC); CSC contracted out the master file replacement software work to IBM. By this summer, however, CSC sources indicated that IBM might not be around when the IRS finally began work on the second phase of the new data engine -- and IRS sources indicated that CSC itself might not be around much longer. (For prior coverage, see 2003 TNT 166-3 or Doc 2003-19286 (4 original pages).)

On October 17 neither Dalrymple nor Charles A. Lacijan, the IRS Oversight Board's staff director, commented on what the status of the prime contractor will be in the upcoming IRS announcement about the technology modernization effort. However, according to Dalrymple, every group studying some aspect of the IRS's modernization effort has told the agency flat out, "Do not start over."

"It would be easy for me to blame everything on the prime [contractor], but that's just not the case," Dalrymple said.

There are at least four major outside groups studying the IRS's and the prime contractor's progress and problems: the Software Engineering Institute through Carnegie Mellon, Acquisition Solutions, the management consulting firm Bain & Co., and the Gardner Group technology consultant firm. The IRS itself also is analyzing what caused each cost overrun and delay. All of these studies and their recommendations for improvement are either completed or near completion, Dalrymple said.

According to Dalrymple, the preliminary findings of the studies are that:

o not everything is broken in this program;

o the prime contractor underestimated the scope of some of the
IRS's projects; and

o the IRS didn't take on all responsibilities that it should
have, assuming the prime contractor would.

"We underestimated their capacity to overestimate their capacity," Dalrymple said of the IRS and its prime. "And we overestimated our own capacity."

Dalrymple said one of the things the IRS and the prime contractor are finding out through the studies is that the average "big" technology project in the private sector has a $ 5 million budget -- something that would be so small in the context of the IRS's effort that it wouldn't even register as a "tier-B" project at the Service. The success rate of the "big" $ 5 million technology projects in the private sector is about 30 percent, and on average the private sector's costs for those projects is about double the original estimate.

"It's not an excuse but meant to put what we're doing into perspective," Dalrymple said.

The IRS has been told by all the groups studying the modernization effort that they've seen continued improvement. However, the IRS has to realign its priorities -- that is, select the key projects it must implement and focus on getting those up and running. The agency also must clarify roles and the decisionmaking process in the project's management structure.