• E-mail this page
• |  Print this page
• | 
• | 
• | 

How To Rescue Federal IT

In addition to the high-profile failures of past projects like the FBI's Virtual Case System, recent weeks have seen a 72-page report from the GAO about the DoD's continued problems with its business-systems modernization. Among the problems discussed in the report is that the latest version of the business enterprise architecture, version 3.0, doesn't entirely satisfy Congress' requirements. Turns out that version 3.0 includes the target architecture that the DoD is trying to achieve, but fails to include the current systems base. The result is the lack of a comprehensive transition plan to get there from here.

Another link in the most recent chain of federal IT mistakes comes from Tom Ridge, former secretary of the Department of Homeland Security. He recently talked about how IT problems are barriers to that agency's fulfilling its mission, and how a big part of the problem is the federal government's decentralized approach to IT management and spending.

Indeed. As my colleague Eric Chabrow points out in his excellent feature story on the topic, the huge, all-singing, all-dancing federal IT projects are almost set up to fail from the very beginning.

The way the system is supposed to work is designed to allow citizens' elected representatives--that is, Congress--to set budgets and priorities. And in great checks-and-balances fashion, other sets of eyes from oversight agencies and elsewhere are looking to make sure our money is spent appropriately and correctly.

It all sounds great in theory. But what winds up happening with the largest IT projects is that once the money is appropriated, like a giant snowball on its way down the slope, there's no stopping it. In the private sector, when an IT project is going south it's either fixed quickly or the plug is pulled before scads of money and time are spent on it. In the federal government, there's a disincentive for doing this, because not only are mistakes out there for the world to see, but if the money goes away it usually doesn't come back. Money is allocated to specific projects, not to IT in general. If it's not spent on that specific project, the money is gone.

A possible solution to this is offered by Steve Cooper, former CIO at the Department of Homeland Security, in a podcast well-worth checking out. While still working in the government, he proposed allowing agency heads to have a discretionary fund, to allow them to pull the plug on failed systems and divert the money to other things, with appropriate rules and oversight, of course. In other words, he suggested that federal CIOs be allowed and encouraged to actively manage their agency's IT investment, just like executives in the private sector do, without having to ask permission of five different committees any time they wanted to do something like kill a project because it wasn't working. Mr. Cooper's the first to admit that this idea didn't get very far, mostly because people in the Washington power corridors considered it a 'slush fund.' It's all in how you look at it, I guess. But you can hear his suggestion, as well as Eric Chabrow's take on how to administer it, here.

This issue--of allowing federal CIOs to more actively manage their agency's IT investments--certainly looms large. And I'm certainly in no position to disagree with Mr. Cooper here, but I believe that his suggestion would fix only part of the problem. The other big issue is that of culture.

Charles Rossotti, the former IRS commissioner and the first non-tax guy to be appointed to the position, emphasizes the importance of cultural change in how the IRS turned a big, ugly, expensive infrastructure on the verge of collapse into one that's much more efficient. (On a related note, there's a podcast interview with Richard Spires, in charge of the agency's current business-systems modernization plan, that you can check out here.)

As Mr. Rossotti and his team proved with the mostly successful turnaround of the IRS, sometimes you have to break something to be able to fix it. I heard Mr. Rossotti give a keynote speech at a conference a few months back. Although he of course talked about the technology and how it wasn't working, he spent a fair amount of time discussing other issues. One of his early actions, in fact, was to lead a field trip to a private company with what he considered to be excellent service and great technology--in other words, to show top IRS leaders the potential instead of just talking about it. As he said, "We were out to change the way the IRS worked, and not just change the technology."

He also talked about how important it is to help everyone understand that things can and will change. "One big difference is that the sense of pessimism isn't there any more," Rossotti said. To get there, it was critical to "encourage a process of acknowledging when things don't go well"--to be willing and able to "face up to bad news."

Everyone likes a winner, but sometimes you need to discuss the elephant in the living room to figure out how to make it go away.

Which leads to turning the question to you--the real experts--the career IT professionals who work for the federal government: What now? What needs to be changed, and how? I invite you to participate in the discussion.

« Daily News Podcast, Tuesday, November 29 | Main | Robert Abbe at Jeffries Broadview Talks Up IT Trends »