Every year, DoD is losing flexibility to put money where it matters most. DoD submits a budget with more than 1,800 program elements (PEs) for RDT&E and Procurement. It is supported by more than 10,000 lines of project details across 30,000 pages of documentation.
If an activity of reasonable size is not justified in those 10,000 lines, then it is a new start and requires prior approval from four congressional committees.
If a PM wants to move resources from one of those 1,800 PEs to another by more than $10M or 20% (whichever is *less*), then the PM must ask “Mother may I?” up the chain and to congress. This harms the responsiveness of DoD acquisition to emerging tech, new threats, and fact of life changes like inflation or supply chain delays.
Reprogramming over the last several years, both Above Threshold (requiring congressional prior approval) and Below Threshold (within DoD control), are shown below for each of the four major components. Note the declining trend since the height of the Global War on Terror.
In FY 1961, reprogramming was 20% of the entire RDT&E appropriation. In FY 2020, it was only 2%. Reprogramming that DoD can do “below threshold” added just 1% more flex. This puts a premium on exquisite prediction, not a bias for action.
The reprogramming thresholds of $10M or 20% have not changed in decades despite inflation. In fact, Procurement’s threshold was recently decreased from $20M to $10M in FY 2020. And in most cases, for PEs $50M and below, it is the 20% that limits cumulative reprogramming — not the $10M. The total effective BTRs the DoD components achieve was as high as 4 percent in the early 2010s, declining to roughly 1 percent a year by the end of the decade.
The addition of 20%, whichever is less, to the dollar threshold was only added in FY 2005, written into a conference report for the appropriations act. Again, reprogramming isn’t a law (unless it goes across approp accounts past GTA threshold) — it’s a custom.
Here are more restrictions over time:
- 1961: control moved down from budget activity level to PE level
- 2000: thresholds changed from congressional notification to prior approval
This was from a two-pager on reprogramming trends, and you can read the entire report Execution Flexibility and Bridging the Valley of Death here. #MasonGovCon
Webinar with Leaders in Acquisition
Be sure to join me for a webinar with an all-star panel to discuss PPBE reform, the valley of death, and more! We will have a tremendous panel to discuss this topic including: Christopher O’Donnell, the Acting Principal Deputy Assistant Secretary of Defense for Acquisition; Katie Wheelbarger, Vice President of Global Program Support at Lockheed Martin; and Elaine McCusker, Senior Fellow at the American Enterprise Institute and former Acting Undersecretary of Defense (Comptroller).
Register here!
Having been there/done that, reprogramming actions are a pain in the a**. It requires approval through multiple levels above the PM: PEO, Service Executive, Service Comptroller, DoD AE/Comptroller, etc., etc., all the way to Congress for an ATR. Most PMs basically don’t bother, because by the time the approval comes, if it comes, it is obsolete. Hopefully they’ll (Congress) allow more flexibility given the “speed of relevance” challenge we currently face. But you should recognize that the service chain is as much as fault in delaying things as is the DoD/Congress chain.
No doubt you are correct! I wonder how much of that is due to inflation of busy staff work, and how much is due to the fact if it is going to congress then everyone’s eyes has to be on it for approval.