Restructuring the DoD’s $850B annual budget into broad capability areas to manage RDT&E, Procurement, and O&M via dynamic portfolios to acquire, deliver, and operate integrated suites of capabilities.
Where, in this new structure, are critical issues like interoperability, cybersecurity, data distribution etc. addressed? Even AI. It is remarkably inefficient to consider these issues within service / component / capability area? What am I missing here?
Andrew - Within each capability area, there would/should be funding focused on interoperability, data, and cyber. It would pivot from the paradigm that dozens or hundreds of individual programs manage it for their systems asynchronously. This could be investing in a common platform, standards, interfaces, or suite of capabilities that many capabilities (aka programs) within the portfolio could leverage. Doing this across capability areas is an increased challenge that requires significant work, but should be easier to connect portfolio areas to address kill chains or common elements across Services (e.g. C4I) than hundreds of programs trying to do that today.
Pete - thank for the clarification. I agree this is simpler and better than it was but it seems to me that there should be a cross capability overlay that looks at the things that must work across capability. Like interoperability. I'm experiencing some of this now. UAV'a are acquired based on individual mission suitability. Fair enough. The OEM's acknowledge the interoperability and cyber issues but aren't individually motivated to address them. They are focused on their core product (hw / sw) and how things work together is a secondary consideration. Standards, e.g. MOSA are a backstop and an overhead for OEM's.. I guess my point is this - Ukraine has changed the game, the Houthi's perhaps even more so. Whoever gets all this disparate weapons systems (including autonomous) to work together most effectively will win (see hellscape). That reality should be a cornerstone of the new acquisition approach and I don't think it is.
Great article... seems to ignore the problem of "we can't cut that... 600 people in my district work on that". Congress won't give DOD that level of flexibility because they could decide to build System X (in Wyoming) instead of System Y (in California). That's not acceptable to Senators from California who made lots of deals to get the System Y built in CA.
Great article. I would say that the more that budgetary authority is decentralized, the greater need there is for centralized records. DoD (and fedgov) has struggled with central software solutions, so the better approach would probably be a uniform data taxonomy. Projects could be required to keep records according to this taxonomy.
What sort of records are needed? Accounting information on obligations and payments, ideally in real time. Granular information on projects, as Matthias says. Probably a lot else, too!
Great article and totally agree with the importance of these changes! Two additional recommendations I would add to the great ones you made:
1. Budgets should be published in a public-facing data tool rather than a series of books. It will be important for the unclassified data to still be available publicly (the unclassified only of course) so that not only Congress can view it, but so that industry, academia, and the general public can use the data to build even better tools and decision aids. Ideally, this structure could be expanded beyond DOD as well, with important and related investments in other agencies able to more easily share information and coordinate inter-agency activity.
2. In a new budget structure, a certain amount of descriptive granularity is helpful, at a project-like level. I completely agree that these buckets should be larger and not account for every $5M increment, but there is a lot of value in being able to find common threads of cross-cutting technologies and missions (i.e., artificial intelligence, semiconductor, or CBRN investments) that would not be consolidated into a single capability area. These descriptive projects would have to be set up to benefit from the reprogramming flexibility you outline (below the line of accounting) but still offer some analytical insights. I think having them nested makes more sense than parceling each technology out as a distinct S&T line, as that may only exacerbate the BA3 -> 4 hand-off challenges. Here, I would see the BA distinctions going away, but having non-hierarchical fora focus on those cross-cutting technologies, fostering collaboration that relies upon the insights to find said common ground in these budgets.
Matthias - re #2, you could still capture that in Major Thrusts without losing the flexibility and still have them tied to more strategic goals without losing the fidelity. I definitely think we could find a middle ground there.
Matthias - Fully agree on public data tool and tactical tagging/indexing. Need to revisit the accounting elements for insight and analytics vs reprogramming constraints. As for your BA point, agreed - see our related post on DoD Needs Fewer Colors of Money https://defenseacquisition.substack.com/p/dod-needs-fewer-colors-of-money
Seeing the "Xerox" in the real world (outside of the PDF itself) gave me a chuckle like few have lately
Where, in this new structure, are critical issues like interoperability, cybersecurity, data distribution etc. addressed? Even AI. It is remarkably inefficient to consider these issues within service / component / capability area? What am I missing here?
Andrew - Within each capability area, there would/should be funding focused on interoperability, data, and cyber. It would pivot from the paradigm that dozens or hundreds of individual programs manage it for their systems asynchronously. This could be investing in a common platform, standards, interfaces, or suite of capabilities that many capabilities (aka programs) within the portfolio could leverage. Doing this across capability areas is an increased challenge that requires significant work, but should be easier to connect portfolio areas to address kill chains or common elements across Services (e.g. C4I) than hundreds of programs trying to do that today.
Pete - thank for the clarification. I agree this is simpler and better than it was but it seems to me that there should be a cross capability overlay that looks at the things that must work across capability. Like interoperability. I'm experiencing some of this now. UAV'a are acquired based on individual mission suitability. Fair enough. The OEM's acknowledge the interoperability and cyber issues but aren't individually motivated to address them. They are focused on their core product (hw / sw) and how things work together is a secondary consideration. Standards, e.g. MOSA are a backstop and an overhead for OEM's.. I guess my point is this - Ukraine has changed the game, the Houthi's perhaps even more so. Whoever gets all this disparate weapons systems (including autonomous) to work together most effectively will win (see hellscape). That reality should be a cornerstone of the new acquisition approach and I don't think it is.
Excellent. Great to see such valuable information to read and to process. Thank you.
PPBE must be eradicated.
Start fresh.
Trust? Beyond your scope.
After the new government consolidates constitutional power, and a couple of elections perhaps?
Start over fresh, once power is secured. Move it all to a different and smaller, younger location.
Great article... seems to ignore the problem of "we can't cut that... 600 people in my district work on that". Congress won't give DOD that level of flexibility because they could decide to build System X (in Wyoming) instead of System Y (in California). That's not acceptable to Senators from California who made lots of deals to get the System Y built in CA.
Great article. I would say that the more that budgetary authority is decentralized, the greater need there is for centralized records. DoD (and fedgov) has struggled with central software solutions, so the better approach would probably be a uniform data taxonomy. Projects could be required to keep records according to this taxonomy.
What sort of records are needed? Accounting information on obligations and payments, ideally in real time. Granular information on projects, as Matthias says. Probably a lot else, too!
Great article and totally agree with the importance of these changes! Two additional recommendations I would add to the great ones you made:
1. Budgets should be published in a public-facing data tool rather than a series of books. It will be important for the unclassified data to still be available publicly (the unclassified only of course) so that not only Congress can view it, but so that industry, academia, and the general public can use the data to build even better tools and decision aids. Ideally, this structure could be expanded beyond DOD as well, with important and related investments in other agencies able to more easily share information and coordinate inter-agency activity.
2. In a new budget structure, a certain amount of descriptive granularity is helpful, at a project-like level. I completely agree that these buckets should be larger and not account for every $5M increment, but there is a lot of value in being able to find common threads of cross-cutting technologies and missions (i.e., artificial intelligence, semiconductor, or CBRN investments) that would not be consolidated into a single capability area. These descriptive projects would have to be set up to benefit from the reprogramming flexibility you outline (below the line of accounting) but still offer some analytical insights. I think having them nested makes more sense than parceling each technology out as a distinct S&T line, as that may only exacerbate the BA3 -> 4 hand-off challenges. Here, I would see the BA distinctions going away, but having non-hierarchical fora focus on those cross-cutting technologies, fostering collaboration that relies upon the insights to find said common ground in these budgets.
Matthias - re #2, you could still capture that in Major Thrusts without losing the flexibility and still have them tied to more strategic goals without losing the fidelity. I definitely think we could find a middle ground there.
Matthias - Fully agree on public data tool and tactical tagging/indexing. Need to revisit the accounting elements for insight and analytics vs reprogramming constraints. As for your BA point, agreed - see our related post on DoD Needs Fewer Colors of Money https://defenseacquisition.substack.com/p/dod-needs-fewer-colors-of-money