The fact that Vivek Kundra has decided to leave the federal government to accept a fellowship at Harvard should surprise no one. There are several reasons why it should have been expected, even predicted, despite the fact that Kundra probably didn’t realize it himself initially.
First, trying to implement a huge but vague mandate such as the one he was given is difficult to the point of impossibility. Second, executives placed into these high-profile positions are used to getting results. But in the federal government, results are very slow to come even in the best of times.
Playing into the probable frustration that Kundra must have experienced is the temporary nature of the job. He served at the pleasure of the president, and President Barack Obama’s first term is starting to wind down with a tough re-election campaign fast approaching. Kundra had no certainty that his tenure-and thus the organizational changes he wanted to make-would survive the election.
Then of course, there’s the nature of the federal bureaucracy. The federal CIO has no authority outside the Office of Management and Budget. While he can evangelize the ideas he had aboutopen-source andcloud-based computing, Kundra had no means by which he could actually require such changes. He also had no budget to make them happen. In the Executive Branch, each agency or department runs its own IT. The CIO in each agency reports to the agency head, not to the federal CIO.
This means, ultimately, thatIT managers and agency CIOs that don’t agree with the federal CIO’s direction on how their operation should be run can simply ignore him. They know that they’ll still be around long after the political appointee leaves. So generally, these managers give the ideas lip service, but otherwise they simply out-wait them. This is partly because they know that someone in Washington has no idea what their agency’s data center is doing and partly because they know that the CIO’s suggested changes don’t come with a budget attached, so they’re almost impossible to implement.
I saw how this happened during my years as the executive officer of a large military data center before I retired from the Navy. The managers had been in place for years, and they weren’t particularly interested in ideas that some young officer fresh out of the fleet might have about running data centers. Furthermore, they didn’t have the means to make changes while also doing the jobs they were paid to do.
Kundra Frustrated by Lack of Authority
The only reason I was able to make any progress at all was that I was at a very low level in the federal IT structure, so I worked with these managers daily. I was able to convince them to try one tiny thing at a time. The federal CIO doesn’t have even that small advantage. He works remotely, passes out broad policy objectives, and has no direct contact with people in the trenches.
Knowing all of this, one has to wonder why Kundra was willing to take the job. To understand his motivations, it helps to know a little about how things work here in Washington. First, when the president asks you to take on a job, it’s pretty hard to say no. It’s especially hard when it’s in your area of specialty, you have firm ideas about how things can be improved, and when you know that if you’re successful you can make a huge positive change in how the government works.
Second, Kundra was the CIO of the city of Washington, D.C., when he was tapped for the federal job by the president. Washington probably has the most dysfunctional government in the U.S. It’s easy to suspect that once Kundra found out just how awful things were in the D.C. government, he was already looking for an escape route. I know I would have been.
So once Kundra leaves Washington for Harvard, where does this leave his initiatives? Right now, that’s unclear. Kundra just released a 25-point plan for improving federal IT. There was general agency buy-in. But how much of that buy-in was lip service is unclear. His ideas about allowing the use of consumer products in the federal government may get some traction if only because federal budgets are so tight that CIOs will do anything that might save money. The idea of moving mission-critical applications such as email to the cloud is less clear.
The next person appointed to be the federal CIO will have to work with the same lack of authority and a budget to carry out real changes. In addition, the next CIO will have a host of new challenges, including a very significant level of security threats, antiquated IT environments that are much too old to move to the cloud (or even to new computers) and for which there are no funds for updating those environments. There are also significant operational challenges with many of the ideas that have been floated so far. The bottom line is that the job of federal CIO is a tough one-perhaps too tough for the position as currently envisioned.