Serial Local Optimization Failure¶
Core Idea¶
Serial Local Optimization Failure occurs when stages arranged in an ordered chain each optimize a locally scoped objective, while the effect of each choice on other stages falls outside that local objective. Every stage can make a rational, competent decision and the chain can still produce a result strictly worse than a jointly optimized version of the same system. The distinctive mechanism is compounding: one stage changes the conditions inherited by the next, the next optimizes against that already-distorted condition, and the aggregate deviation grows along the chain.
The abstraction is not simply “local versus global.” It requires a serial handoff. Parallel departments pursuing conflicting metrics may exhibit goal misalignment, but they do not instantiate this prime unless their decisions enter one another's conditions in sequence. It is also not a local optimum on a rugged landscape: no stage must be trapped. Each may be at its true optimum for the objective it was given.
Structural Signature¶
- an ordered chain of two or more decision stages
- a locally scoped objective and control variable at each stage
- a handoff through which one stage's choice changes another's feasible set, cost, price, load, or payoff base
- exclusion of that interstage effect from the choosing stage's objective
- individually rational local choices
- compounding deviation across stages
- a joint objective or integrated benchmark that yields a better aggregate result
- a coordinating, contracting, repricing, or integration mechanism capable of internalizing the cross-stage effect
What It Is Not¶
It is not every coordination failure. Equilibrium-selection problems concern which common state agents choose; this prime can have a unique, perfectly predictable outcome that is nevertheless inefficient.
It is not local_optimum. A local optimum is a point best within a neighborhood but not globally best. Here each stage may solve its own problem globally; the loss comes from incompatible objective boundaries across serial stages.
It is not price_of_anarchy. Price of Anarchy measures an equilibrium-to-optimum ratio over a class of games. Serial Local Optimization Failure names one mechanism that can create such a gap.
It is not necessarily an agency problem. The stages may share an ultimate interest yet lack a rule that makes each local decision internalize its effect on the chain.
Broad Use¶
The pattern appears in vertical supply chains where successive firms stack markups, cascading turnover taxes where each levy applies to a tax-inflated base, multi-hop networks where each carrier prices or routes locally, and multi-tier service systems where each layer adds buffers or overhead without bearing the end-to-end delay. It also appears in administrative chains whose units minimize their own visible risk by passing delay, documentation, or cost to the next unit.
Clarity¶
The diagnostic is: If every stage is making its locally best choice, where does the aggregate loss enter? If the answer is “through an effect that crosses the handoff but is absent from the chooser's objective,” the problem is not incompetence. It is the partition of the objective across an ordered chain.
Manages Complexity¶
The prime compresses many “stacking” failures into one audit: map the stages, name each local objective, trace what each choice changes for the next stages, and compare the decentralized result with a joint benchmark. Remedies then sort into a small family: internalize the cross-stage effect, redesign transfer prices, unify control, share a chain-level objective, or add a feedback contract.
Abstract Reasoning¶
For stages \(i=1,\ldots,n\), let \(x_i\) be the choice of stage \(i\), \(u_i(x_i;x_{-i})\) its local objective, and \(W(x_1,\ldots,x_n)\) the chain-level objective. The decentralized solution satisfies each local optimum condition, yet may differ from
The gap persists when \(\partial u_i/\partial x_i\) omits effects of \(x_i\) on other stages' payoffs or on the shared throughput base. Serial coupling makes those omissions compound rather than merely add.
Knowledge Transfer¶
Transfer requires mapping five roles: ordered stages, local objectives, interstage coupling, an omitted cross-stage effect, and a better joint benchmark. The vocabulary of markup, tax, packet, approval, or overhead can disappear while those roles remain.
Examples¶
A three-hop service chain lets each provider add a fee that maximizes its own revenue. Each provider treats the reduced downstream volume as someone else's concern. The final price suppresses total use enough that all providers earn less than under a jointly chosen fee schedule.
Structural Tensions¶
T1: Autonomy versus chain performance. Local discretion uses local information but can exclude remote effects. Diagnostic: Which decision rights can remain local while the cross-stage term enters the objective?
T2: Coordination gain versus coordination cost. Joint optimization can improve the outcome but impose information, bargaining, and control overhead. Diagnostic: Is the recoverable gap larger than the cost of internalizing it?
T3: Efficiency versus distribution. A joint solution can enlarge total value while changing which stage captures it. Diagnostic: Does resistance concern efficiency or the division of the recovered surplus?
Substrate Independence¶
Strip away firms, taxes, routers, and agencies. What remains is an ordered chain of locally optimizing stages, a cross-stage effect excluded from each local objective, compounding deviation, and a better joint solution. That structure is the prime.
Relationships to Other Abstractions¶
Current abstraction Serial Local Optimization Failure Prime
Parents (2) — more general patterns this builds on
-
Serial Local Optimization Failure is part of Optimization Prime
A serial local optimization failure contains optimization because every stage selects what is best for its own scoped objective rather than making an arbitrary or mistaken choice.
-
Serial Local Optimization Failure is part of Pipeline Prime
A serial local optimization failure contains a pipeline because its locally optimizing stages must be arranged in an ordered chain whose outputs or decisions become the next stage's conditions.
Children (1) — more specific cases that build on this
-
Double Marginalization Domain-specific is a kind of Serial Local Optimization Failure
Double Marginalization is the vertical-pricing species of Serial Local Optimization Failure in which firms in a production or distribution chain independently add markups against a shared final-demand curve.
Hierarchy paths (4) — routes to 3 parentless roots
- Serial Local Optimization Failure → Optimization
- Serial Local Optimization Failure → Pipeline → Decomposition
- Serial Local Optimization Failure → Pipeline → Iteration
- Serial Local Optimization Failure → Pipeline → Modularity → Decomposition
Distinction from Neighbors¶
pipeline supplies the serial arrangement and optimization supplies the locally rational choices; neither alone predicts failure. externality is an exact constituent of many priced instances, including Double Marginalization, but the corpus defines it through a price-system frame and it is therefore not required of every substrate-neutral instance. goal_congruence_alignment concerns aligning objectives generally; this prime adds ordered handoffs and compounding. social_dilemma requires a particular strategic payoff structure; this prime does not.
References¶
Citation leads for the Claude re-authoring pass: work on double marginalization and vertical integration; team-theory and decentralized control; supply-chain coordination; cascading turnover taxation; and distributed routing. Independently verify all sources before creating FACT anchors.