Are MDAPs Too Big to Succeed?

Applying lessons from Bent Flyvbjerg to DoD's major programs

The Wall Street Journal had a thoughtful article: 99% of Big Projects Fail. His Fix Starts With Legos.

University of Oxford economist Bent Flyvbjerg is an expert in the planning and management of “megaprojects,” his name for huge efforts that require at least $1 billion of investment.

He’s spent decades wrapping his mind around the many ways megaprojects go wrong and the few ways to get them right, and he summarizes what he’s learned from his research and real-world experience in a new book called “How Big Things Get Done.” 

His seminal work on big projects can be distilled into three pitiful numbers: 

• 47.9% are delivered on budget. 
• 8.5% are delivered on budget and on time. 
• 0.5% are delivered on budget, on time and with the projected benefits. 

“You shouldn’t expect that they will go bad,” he says.
“You should expect that quite a large percentage will go disastrously bad.” 

The DoD’s version of the megaproject is the MDAP - Major Defense Acquisition Program. These are acquisition programs that will cost at least $525M in RDT&E (development) or $3B in Procurement (production).

As the DoD and GAO regularly report, MDAPs have a track record comparable to Flyvbjerg’s megaprojects. Note the top 10 failed defense programs of the current era.

In studying the 0.5% success stories, Flyvbjerg found two lessons:

1. Think slow, act fast.

“The irony of megaprojects is that many are late because not enough time is spent planning.”

The DoD arguably spends too much time planning. It will spend years defining and coordinating requirements validated by a board chaired by the VCJCS, analyzing alternatives, breaking down the program into a work breakdown schedule with an integrated master plan and schedule, a cost estimate and independent cost estimates by the Service cost organizations and CAPE, along with dozens of program strategies reviewed by dozens of oversight organizations. Congress plays an active oversight role before appropriating any program funding. MDAPs often spend over five years from initiation before starting development.

Sure, some programs rush through planning, which oversight is quick to criticize. Others will have overly optimistic cost estimates to gain approval in a tight budget cycle, with everyone buying into the lie - the program office, Service leaders, OSD, Congress, and the prime contractor.

Requirements documents, even after two years of development and validation may include overly optimistic performance parameters. They may require three technological miracles to occur during development, instead of basing a program off mature technologies and iterative updates.

These programs have acquisition program baselines that lock down cost, schedule, and performance and track progress against these baselines as the primary measure of success or failure.

The long schedules for planning, development, and production provide ample time for changes to operations, threats, technologies, budgets, and businesses to drive critical program risk. Most MDAP contracts are cost plus. This means if the contractor slips schedule, they get to charge more to the DoD.

Current rapid acquisition programs impose a five year constraint to rapidly deliver operational capabilities. They are intended to accelerate learning and pivot when necessary. DoD has already demonstrated success across a variety of platforms.

The secret of the B-21 not falling into the same delays as related aircraft programs? They learned how to limit bureaucracy. The quote of the year so far is:

“There were fewer checkers checking the checkers. Don’t ever underestimate the ability of the Pentagon bureaucracy and these many, many reviews to slow the doggone thing down.”

- Former Secretary of the Air Force Deborah Lee James.

2. Find the Lego that simplifies your work and makes it modular. 

“Modularity is a clunky word for the elegant idea of big things made from small things”

“Modularity is a clunky word for the elegant idea of big things made from small things,” he writes. “Look for it in the world, and you’ll see it everywhere.” Everywhere includes “software, subways, hardware, hotels, office buildings, schools, factories, hospitals, rockets, satellites, cars and app stores,” he writes. “They’re all profoundly modular, built with a basic building block. They can scale up like crazy, getting better, faster, bigger and cheaper as they do.” 

Modular Open Systems Approach (MOSA) has been DoD policy for over a decade. Congress recently required it be the default approach in statute, with waivers to adopt a different model. Yet DoD’s track record on MOSA is bleak. This is due in part to the power of the prime contractors to dictate terms like closed, proprietary solutions.

After decades of DoD policies driving defense prime consolidation where 51 primes merged into 5, there are often only 2-3 primes left to compete in each sector. While there are many mid-size companies and up-and-comers eager to win some major awards, the primes enjoy a competitive advantage and protection from competition.

DoD needs to enable a more modular approach to enable a broader industrial base. This includes tech startups to insert key capabilities, particularly software intensive and AI solutions. DoD needs to work with the prime contractors to pivot the business model. Instead of focusing profits on production and sustainment (where bad designs drive higher support revenue), enable rapid upgrades and Software-as-a-Service (SaaS) models. This will accelerate operational performance for our warfighters to maintain or regain a competitive advantage against near-peer adversaries.

Furthermore, DoD needs to avoid the winner-take-all contracting strategies that will define a defense sector for the next decade. There must be robust competition and a regular cadence of future contracts to control costs and fuel innovation. DoD can also enable modularity by transitioning from monolithic MDAPs to capability portfolios that leverage common platforms, components, and services.

DoD should explore more programs that are below the MDAP levels that can achieve a comparable mission outcome. Programs that value low cost, high quantities, and short lifespans instead of trying to design and sustain MDAPs for 30-50 year lifespans. Programs with shorter lifespans enable competing and iterative solutions to emerge every few years.

As Christian Brose wrote in The Kill Chain:

It requires a sweeping redesign of the American military: from a military built around small numbers of large, expensive, exquisite, heavily manned, and hard - to - replace platforms to a military built around large numbers of smaller, lower - cost, expendable, and highly autonomous machines. Put simply, it should be a military defined less by the strength and quantities of its platforms than by the efficacy, speed, flexibility, adaptability, and overall dynamism of its kill chains.

If DoD were able to pivot more towards mission outcomes as challenges and services, industry can leverage their own R&D and design novel solutions to achieve them. From disrupting space launch to commercial satellite imagery to counter-UAS-as-a-Service, there are recent success stories that can disrupt a broader operational and acquisition model. The Secretary of the Navy pondered at a public event recently a CONOCO — contractor-owned, contractor-operated model with commercial autonomous ships. DoD needs to break from the monolithic MDAP models.

For all those involved in DoD’s major systems, from the program offices, industry, or oversight in the Pentagon and Congress, I strongly recommend Flyvbjerg’s new bestseller: How Big Things Get Done