The Fearful Administrator Explosion

Large Capital Projects are getting worse, not better, despite decades of investment in better tools, tighter controls, and ever more sophisticated best practices. When reviewing the results if thousands of large projects, over 90% overrun on cost, schedule or both.

The standard response to this deterioration is always the same: add more governance and control.

Unfortunately, this actually supports the mechanism of failure. These projects do not primarily fail in execution, they fail in design. They are architected on the assumption that the process itself is of such high quality that individual judgment can be safely outsourced to it. It assumes that sufficient insight and foresight is available to make decisions in advance and neutralize second- and third-order effects in complex, non-linear systems.

In complex initiatives, adding governance doesn’t fix the problem, it is the problem. As initiatives scale, organizations typically migrate from High Performance Systems to Administrator Systems. Administrator Systems optimize for safety and insulation, with shared decisions, procedural accountability, reversible commitments, and governance structures designed to protect people rather than outcomes, while activity increasingly substitutes for progress. Process is all, and all is process.

High Performance Systems are designed differently. Decision ownership is singular, accountability and risk are personal, and governance exists to protect results.

When results are in jeopardy, adding control accelerates the Fearful Administrator Explosion, as exposure disappears, judgment erodes, and motion increases.

This is not a project management problem, but a system design problem.

Organizations can survive bad decisions, but they cannot survive outsourced decision making protected by process.

The solution is not more control, but the deliberate reintroduction of judgment through single owners, irreversible commitments, and visible consequence, because without skin in the game, no system performs.

Boardroom Drill

Where have you improved the process when the real design flaw was removing judgment?