Disclaimer: I have consulted with various US federal government and state agencies in the past, and the views expressed herein are mine alone, and not those of the US or any state government agency.
President-elect Obama has proposed the appointment of a Federal Chief Technology Officer to oversee and overhaul federal IT operations - bringing US government IT into the 21st century with updated technologies, transparency, connectivity, etc. etc.
A great number of industry pundits and bloggers have jumped on the bandwagon with many suggestions that not only encompass what to do, but who could be qualified to do it - Bill Gates, Vint Cerf, Jeff Bezos, etc.
Todd Biske has a number of thoughtful suggestions about what a Fed CTO might accomplish, and my commentary is going to be a riff on his views, with a bit of a difference: I'm positing that a Fed CTO will be rendered largely helpless and ineffective unless various federal laws and regulations that drive and directly affect government IT are overhauled - including broad and mandatory standardization. Unless this happens, and fairly quickly, no amount of technology or infrastructure is going to turn this bloated boat around anytime soon.
The federal government operates under myriad rules and regulations that are often specifically tailored to agency goals and operations. In addition to US statutes passed by Congress and signed into law by presidents, we have the Code of Federal Regulations (CFR), Executive Orders, the decisions of various federal judges/panels, the Tax Code (part of statute, but a monster in its own right), agency/commission orders, procurement rules, OMB compliance mandates, rules, regulation, more rules, and on and on and on....
And I haven't mentioned, at least directly, the Clinger-Cohen Act (a.k.a. The Information Technology Reform Act of 1996) yet. That one mandated CIOs for all government agencies in addition to creation and 'mandatory' utilization of the Federal Enterprise Architecture Framework (FEAF) - which is essentially a Zachman framework adapted specifically for the federal government. The positive impact that Clinger-Cohen has had on government IT operations is highly debatable, and needs to be revisited and perhaps overhauled as well.
Confused enough yet? I don't know anyone that wouldn't be. It's a big mess, and it can be untangled, but it will take time, a great deal of effort, and of course, money...but that's what the Federal Reserve's printing presses are for - given that they're currently rolling along smartly for banks, insurers, credit card companies, and perhaps the automobile industry.
One of the fundamental challenges a Fed CTO faces is that most government agencies are very insular when compared against their sister agencies - they operate their piece of the government largely in a vacuum, and that's primarily due to all of the specific laws and regulations that they must honor and account for. A further conundrum is that a great deal of these laws and regulations are specific to an agency, and not the others. It's no surprise then, that agency processes/procedures, and thus, their IT, mirror this specificity and complexity.
Then there is the issue of procurement. Suffice it to say that complexity and obfuscation rear their respective ugly heads here also, and needs to be overhauled. The focus on procurement issues is not just protecting the taxpayer from getting ripped off (which they often are anyway) but that the government derives value from the purchases - immediately and on-going. This would suggest that procurement needs to be streamlined, incremental, much more nimble to the market and emerging useful technology, and swing away from 'big-bang" projects/deployments that take years, cost much, and often fail.
Lastly, there are the issues of us, the people of this great land. We want transparancy, but do not violate our privacy. We want cost-effectiveness, but responsive and supportive agency functions, when we desire and need it. We value security, not just of the homeland, but of ourselves and our individual histories as represented by the data that the government specifically collects about us.
What the new federal CTO faces is not just a 'technology' or IT problem, nor can it be solved simply by some name-brand IT whiz throwing interoperable, SOA-infused, Web/Enterprise 2.0 technologies at it. We have fundamental problems with the enormous bureaucracy that we have created over the years, and that has to be changed and optimized first before the technology comes in, because just throwing technology at the problem is not going to work.
Finally, I have a viable candidate for the CTO position, even though he is not as well known as some of the heavyweights bandied about by IT pundits and bloggers...how about this guy? I think he'd fit the position well.