A critical debate has been stirring since the Army announced its latest approach that requires the use of cost-type contracts when buying software. This, at first glance, looks rather arcane, but the outcome could be monumental if it prevents the Army from being able to compete in the emerging era of artificial intelligence, data analytics, and autonomy.
The problem is that no one else on the planet develops software using the US government’s innovation-crushing cost-plus acquisition model. Department of Defense leadership and Congress should be asking the simple question of why that is the case before the Army embarks on its latest quixotic quest to potentially waste tens of billions of dollars.
If anyone missed it, software has been “eating the world” since well before Marc Andreessen’s 2011 article stated what is now obvious. American commercial software companies have been leading the globe by reinventing an innovation model that ironically once held sway in US defense circles in the 1950s. In the process, these companies have been able to attract the world’s best engineering talent and, so far, stay above the competition. If you think the US Army would be embracing how best to use this extraordinary national asset to its benefit—well, you would be wrong.
Specifically, the Army wants to mandate developing software through cost contracts as it would if it were buying a tank or helicopter. Never mind that this system, which ignores commercial market incentives and then forces a reliance on government-unique contractors that specialize in complying with government rules and regulations, has rarely succeeded in the troubled, stop-and-start history of the Army’s modernization efforts since the 1990s.
Still, Army acquisition officials seemingly want to push the easy button and continue to do business the way they always have. Traditional defense contractors think this is a wonderful idea as it locks in their government-imposed, expensive, and outdated business practices and ensures continued sales. The few Silicon Valley and other non-traditional firms who want to do business with the government are crying foul as they rightly argue it will keep them out of the defense market and bar the government from getting the best solutions.
At the heart of this issue are the differing incentives inherent in government-unique cost vs. commercial fixed-priced contracting. The traditional military contracting approach is to pay specialized defense companies their costs plus a percentage profit fee to then conform to a set of rigid government requirements to be completed through a linear, waterfall acquisition process that develops military capability in timescales that span decades to deliver anything new to the warfighter. This is policed by a government-unique accounting system and overseen by multiple bureaucracies such as the Defense Contract Audit Agency, Inspector General, and Defense Contract Management Agency.
Cost contracts focus on limiting the percentage of contractor profit margins, but, to the taxpayer’s dismay, not overarching total costs. This disincentivizes speed, cost cutting, and efficiency, as the key to maximizing total profits is to let costs increase, or, at a minimum, spend every dollar allocated to a program. There is also a rational fear that the cost model destroys the innovation culture of a company by focusing it on compliance with non-commercial practices rather than results. For this reason, leading commercial companies have historically refused to take cost contracts, while traditional defense companies have been compelled by the government to specialize in them.
In contrast, the commercial fixed-price model embraced by the US software industry starts with the creation of minimally viable products developed to solve a need. These solutions then progress through multiple iterative, time-based releases of new capabilities that are continuously updated based on user feedback. It is results based, rapid, and agile with outcomes measured in days, not years or decades. Commercial software solutions are developed and sold in fixed price packages, as a service, or as licenses. They often contain underlying intellectual property that a company has previously invested its own money in to quickly solve a customer’s problem. Fixed price contracting encourages innovation and greater efficiency as contractors maximize profits by looking to save money and minimize time to results.
30 years ago, in the Federal Acquisition Streamlining Act of 1994, Congress recognized that the two contracting models were completely different and that the preferred pathway was through fixed-price commercial. Instead, the Army in its new software policy has once again ignored the legal preference for commercial and turned its back on Silicon Valley. It will have to re-learn the lesson that while Silicon Valley does not need the Army’s money under these terms, the Army really needs Silicon Valley to compete against China. This can only happen once the Army changes its acquisition incentives.
The post The Army Thinks It Can Build Software Better than Silicon Valley: Good Luck with That appeared first on American Enterprise Institute – AEI.