The end of TechStat
The diminished frequency of Office of Management and Budget TechStat meetings is a likely indicator of an OMB oversight effort plodding toward an end.
OMB has already publicly acknowledged that it has held fewer TechStat meetings--but in a note included with its response to FierceGovernmentIT's latest round of FOIAing TechStat meeting information, OMB attempts to portray this state of affairs as a success. OMB, the note says, originally intended to shift the TechStat workload "to internal agency governance and management while maintaining OMB involvement in a smaller number of Tech Stat reviews."
Naturally as a result, "the number of OMB-led TechStats decreased from the initial number of reviews conducted at the onset of the program as agencies ramped up their review processes and began conducting more of their own TechStat reviews."
The problem with accepting that is that inherent in OMB's rollout of TechStat was a belief that agencies weren't doing a good job overseeing their information technology systems. Is OMB really saying that it turned around that lack of institutional oversight in less than two years? (How, exactly? By providing them a seating chart template for use in homegrown TechStat meetings?)
Further, OMB sought to buttress the TechStat process through a big dollop of cash; in the fiscal 2012 budget request, OMB had asked for $60 million in a line-item appropriations request to support TechStat meetings. Vivek Kundra, the then-OMB administrator for e-government and IT (aka the federal chief information officer) told a House panel in May 2011 that the difficult part of TechStat isn't the meetings.
"It's actually the follow-through and the follow-up, which takes countless hours and resources to make sure that if Agency A has committed to making sure that they're going live in one month, that we come back a month from then and say, 'You said you would go live, what happened?'" he said.
Had Congress come through with that appropriation, in other words, it's a pretty safe bet to assume that OMB would not be talking about how it had always planned to make TechStat the responsibility of agencies.
In any case, the gradual disappearance of TechStat really doesn't matter. That federal IT is in dire need of major reformation is obvious, and TechStats were just a diversion along the way--and not a very well thought-out one.
For example, a frequency analysis of agencies the subject of all known OMB-led TechStat reviews since OMB began holding the reviews in early 2010 shows the Homeland Security Department as the most recurrent target. Of the 71 known TechStat meetings OMB has called during the past three years, nine, or nearly 13 percent, have been regarding a DHS system. The Defense Department, which spent nearly six times as much as DHS on unclassified and non-embedded information technology systems during fiscal 2011 ($33 billion versus $5.6 billion) was the subject of just eight TechStats, or about 11 percent.
Similarly, Housing and Urban Development, which spent $478.4 million on IT in fiscal 2012, has been the subject of the same number of systems as the departments of Treasury ($3.2 billion) and Transportation ($3 billion).
Those numbers suggest less-than-rigorous cost-benefit analysis when it comes to selecting which systems to evaluate during a TechStat.
Even in the systems that have come under review, it's easy to wonder whether the time it took to subject it to a TechStat was worth the time.
For example, on Nov. 9, 2010, OMB subjected the Federal Aviation Administration's En Route Automation Modernization air traffic control software project to a TechStat--the same system that the Transportation Department office of inspector general said just last September is still subject to critical system failures.
And so on. It's easy to cherry-pick failures from a list of 71 TechStat meetings, but given that agencies have long treated OMB oversight efforts as check-the-box exercises to be borne through as quickly as possible, it's easy to suppose that the net effect of many TechStat meetings is marginal.
In fact, what other conclusion can there be? If TechStats are great at turning around troubled agency projects, it would be inexcusable for OMB to lessen their frequency: Surely it wouldn't stop its own oversight effort had it demonstrated a spectacular success rate, big appropriations or not. If, on the other hand, TechStats are indeed marginal--well, that explains why there have only been 11 in the past 2 years. - Dave