Feedback on federal contracting

Writing over the weekend about Vivek Kundra, I said that if Obama planned to seriously make government more efficient, he would need to reform contracting:

[Without reform], government IT will never match the private sector’s and more importantly we will continue to waste taxpayer money at a time when we can least afford to do so. If reform succeeds it will benefit government and taxpayers for decades to come.

I received two interesting comments that pushed back on that sentiment.

As a former USAF acquisition person -- currently teaching government contracting -- my response is "good luck." The Federal Acquisition Regulations are passed by Congress (not the Executive). They drive average acquisition lead times to almost two years in order to get through all the hoops. Unless Congress is willing to step to the plate and reduce some of the requirements for small business and other set-asides; to shorten the time for competitive responses; and lots more -- we are not going to get anywhere. By the way, no-bid contracts are a thing of war as far as DOD is concerned. A no-bid contract is almost unheard of except in emergencies.

I don't really have an opinion on whether Congress will be able to do this. I admit it seems unlikely but I would suggest that the ONLY way it could happen would be for the Administration to make the push and to push hard for Congress to reform the rules. I think this is an area that will fit under the "a crisis is a terrible thing to waste" umbrella. The financial crisis is obligating the government to spend huge amounts of money, Republicans are asking for fiscal discipline (which is of course the worst possible thing to do in a deep recession but Obama is promising line-by-line scrutiny to cut unnecessary costs). I have no doubt that billions can be saved by allowing the government to operate more freely, by contracting with less red tape and more accountability -- that is, as any rational business would act.

Unfortunately, another commenter wrote, it's not that simple.

I think it's important to recognize that Government is (and should be) held to a different set of standards than private business. Government is required by regulation to deal equitably with all willing contractors. Private business, on the other hand, is under no obligation to offer equal opportunities to all interested offerors and can work within existing business relationships, even if the contract cost is greater. This probably accounts for business's greater agility.

Well ... in business, you are always trying to cut costs, to get vendors to compete with one another. But you also realize that the cost of a purchase is not just the amount on the bottom line, it's also the cost of maintenance contracts, the associated internal costs, the training costs. Most importantly, it has to work and you have to calculate the lost productivity, the lost hours when something doesn't work as advertised. You build relationships with trusted vendors and you don't risk big projects on unproven vendors because they bid 10 percent cheaper. The problem with government IT is that there is a tendency, even a need to build huge, mammoth systems that are so complex they are inherently primed for failure. Everyone knows the projects will fail but neither the vendor nor the government employees are able or interested in stopping them. What I think Kundra can offer is a completely transformed mindset away from giant projects and towards breaking IT into smaller bits that can talk seamlessly and securely with one another. I don't know exactly what that looks like but I think that's going to be how you get the right level of data- and knowledge-sharing. Since the big vendors have failed so spectacularly, I think the door is wide-open for agencies to put real programmers on its payroll to create projects like the ones Kundra has created, doing it with open interfaces so that the open source community can create new applications, and letting small fast-moving developers into the bidding process. Perhaps this is all blue sky and the realistic response is Good Luck. Or perhaps something hugely better is within our grasp. Government IT is not the problem. Bad IT (and in general terms, IT contractors) is the problem.