Governments tend to grow because complex systems accumulate fixes faster than they shed them; risk aversion and incentives favor adding layers over removing them; and chasing “just a bit more performance/coverage/safety” drives exponential cost and bulk. That’s pure Murphy, classic Systemantics, and textbook Augustine working together as a one‑way ratchet [1][2][3].
How the growth ratchet works
- Failure-driven accretion: Every visible mishap yields at least one new rule, office, or report; almost no one gets rewarded for removing the old ones. Adding is cheap politically; subtraction is risky. So the rulebook only grows [1].
- System self‑preservation: Once created, a system devotes increasing energy to keeping itself alive. Outputs (reports, metrics, “findings”) become inputs that justify more resources, controls, and coordination layers. The organization gradually optimizes for itself rather than its original mission [2].
- Chasing the last 10%: Pushing toward zero risk, total coverage, or perfect compliance produces exponential cost and organizational bulk. Each “tighten the spec” step demands more oversight, more interfaces, and more staff to manage the added complexity [3].
- Information distortion and coordination tax: Information gets filtered up the hierarchy, so leaders add more reporting to “see clearly,” which spawns compliance units to feed the reports, which then need coordinators to reconcile contradictions. Each fix creates the need for another fix [2][1].
- Budget and constituency lock‑in: “Use it or lose it” budgeting and the creation of beneficiaries make programs sticky. Concentrated gains + diffuse costs = growth now, pruning never. It’s easier to add a new program than to kill an old one [1][3].
- Crisis ratchet: Emergencies justify temporary expansions that rarely roll all the way back. Temporary turns out to be the most permanent word in government [1][2][3].
The laws involved, in plain language
- Murphy’s Laws (institutionalized): If a regulation can expand to cover an edge case, it will; side effects will arrive on schedule; and any fix big enough to work will create problems big enough to need another fix [1].
- Systemantics (Gall’s insights): Complex systems that “work” are evolved by piling on patches, not redesigned cleanly; systems naturally drift from serving their missions to serving their own survival; and adding subsystems usually creates more failure modes than it removes [2].
- Augustine’s Laws (of large programs): The last increments of performance cost the most; complexity and cost rise nonlinearly with requirements; and schedule pressure begets oversight begets personnel growth—while delivering diminishing returns [3].
Why it seems like “always”
- Asymmetric friction: It’s easy to add one more layer to avoid blame; it’s hard to remove one and assume risk. Over time, the small additive decisions dominate the rare subtractive ones [2][1].
- Path dependence: Each new layer creates stakeholders, metrics, and interfaces that make reversal costlier than compliance. Yesterday’s workaround becomes today’s essential function [2][3].
If you want the growth to slow, you have to reverse the ratchet: make subtraction safer than addition (sunsets with proof‑of‑value to renew), cap layers, reward risk‑balanced simplification, and budget for outcomes rather than activity. Otherwise, Murphy supplies the events, Systemantics supplies the mechanics, and Augustine supplies the bill—on time and over budget [1][2][3].
Sources
1
Augustine's Laws by Norman R. Augustine
Systemantics by John Gall (not systematics)
3
Murphy's Laws by Arthur Bloch
In addition:
Here are additional angles that clarify why government size drifts upward and which “laws” make the drift feel inevitable.
Deeper mechanics of the growth ratchet
- Interface multiplication: Every new program introduces interfaces to existing ones; coordination overhead grows faster than headcount, which then justifies more coordinators and rules to manage the new interfaces [2].
- Accountability conservation: When the chance of blame rises, systems add oversight layers because adding looks safer than simplifying; each layer adds failure modes and delays, which in turn “require” more oversight [1].
- Policy sedimentation: Crises create fast patches; peacetime lacks symmetrical energy to remove them, so yesterday’s workaround becomes today’s sacred requirement and tomorrow’s staffing line item [3].
- Goodhart drift, system edition: Once a metric becomes a target, people optimize the metric; leaders respond by adding more metrics and compliance checks, expanding the bureaucracy that services the numbers instead of the mission [2].
- Mission creep as survival strategy: Programs adopt adjacent missions to stay relevant; each “small extension” brings new stakeholders, processes, and budgets that rarely unwind [1].
A typical lifecycle of program bloat
- Birth: A visible failure or crisis triggers a rapid build with broad mandates and thin definitions of “done,” which seeds future control layers [1].
- Adolescence: Measurement arrives to demonstrate value; reporting units and compliance offices stand up to feed the dashboards [2].
- Maturity: Edge cases and exceptions drive a thicket of rules; appeals and waivers add parallel pathways and committees to arbitrate them [2].
- Senescence: The program becomes a self‑licking ice‑cream cone—most energy sustains itself, not the original mission, yet any underperformance is used to argue for more resources [3].
Telltale indicators you’re in the bloat zone
- More people are counting the work than doing the work; the metrics look comprehensive while delivery slows [2].
- Decision latency rises nonlinearly with the number of stakeholders, and nobody can name the single owner empowered to remove a step [1].
- “Temporary” measures from a past emergency are still active, renewed automatically, and now have dedicated staff and training modules [3].
- Budgets operate on “use it or lose it,” so spend spikes at year‑end on low‑value items to protect next year’s baseline [1].
- Exceptions are more numerous than rules, and the appeals process employs more people than the core service line [2].
Why pruning attempts often fail
- Concentrated pain, diffuse benefit: Every cut has a small, organized group to fight it and a large, unorganized public to shrug, so the path of least resistance is addition [1].
- Postmortems that add steps: After failures, the default corrective action is “add a check,” not “remove a step,” because removal feels risky and is hard to credit politically [2].
- Diminishing returns trap: Pushing for the last 10% of coverage or risk reduction explodes cost and oversight while producing marginal gains that are politically irresistible to promise [3].
What actually helps (consistent with the laws)
- Subtraction as the default: Require that every new rule, role, or report sunset unless it proves net value on outcomes, not activity; renewal must be justified as if it were new [1].
- Regulatory and oversight budgets: “Two out, one in” by function, not page count, so interfaces and sign‑offs are capped; exceed the cap and something must be retired first [2].
- Pre‑specified off‑ramps: Time‑boxed authorities with decay functions (automatic scope and resource reductions each period) so “temporary” stops being permanent by inertia [3].
- Kill‑criteria tied to mission metrics: Define in advance the thresholds that trigger shutdown or consolidation; if the metric fails twice, the default is terminate, not expand [1].
- Modularization and parallel pilots: Build small, loosely coupled units and evolve what works rather than bolting fixes onto a monolith; retire failed modules quickly [2].
- Reward removals: Career credit and budget carry‑over for validated deletions and cycle‑time reductions; make subtraction safer than addition for managers [3].
- Pre‑mortems and reverse‑AARs: Before adding a control, ask “what failure will this create elsewhere?” and after incidents, test “what can we remove to reduce complexity next time?” [2].
Metrics that keep the system honest
- Decision lead time from intake to yes/no, not proxies; trend should fall or trigger a rollback review [1].
- Ratio of mission staff to compliance/reporting staff; drifting downward signals self‑preservation over service [2].
- Net rules deleted per quarter and interfaces retired per change; if adds outpace removals consistently, expect exponential overhead growth [2].
- Share of spend on maintenance versus new capability; crossing certain thresholds predicts Augustine‑style cost blowups for tiny performance gains [3].
- Exceptions‑to‑rules ratio; rising exceptions indicate the rule set has outgrown reality and is spawning its own corrective bureaucracy [1].
Common questions, answered tersely
- Is bigger always worse? No, but past a complexity threshold the coordination tax outgrows the benefits of scale, and the system begins serving itself more than the mission [2].
- Will technology fix it? Only if it reduces interfaces and steps; digitizing a bad process just produces bad outcomes faster and adds another subsystem to fail [1].
- Can crises be handled without permanent growth? Yes—if emergency powers and programs are created with hard sunsets, decaying budgets, and pre‑committed rollback playbooks [3].
Bottom line
- Murphy supplies frequent surprises, Systemantics explains why fixes accrete faster than they’re pruned, and Augustine quantifies the runaway cost of chasing perfection; together they form a one‑way ratchet unless you make subtraction safer, interfaces fewer, and renewal genuinely earned [1][2][3].
Sources
1
Augustine's Laws by Norman R. Augustine
2
Murphy's Laws by Arthur Bloch
3
Systemantics by John Gall (not systematics)
No comments:
Post a Comment