Procurement workflow design under a shared legal and governance boundary

Abstract

ProcuraCost compares the declared cost of two lawful procurement workflow designs: formal sequential and adaptive compliant. Both alternatives retain the same legal and governance boundary, procedure family, initiation date and mandatory legal waits. The model therefore tests workflow and contract design without treating compliance as a variable to be relaxed.

The calculation is deterministic. Scenario values are starting assumptions with explicit provenance, not measured organisational effects or probability estimates. The model records the direction and range of the cost difference but does not select a procedure, infer implementation readiness or prescribe an outcome.

Calculation contract

Each alternative has its own directed process map with activities, dependencies, role effort, waiting time and non-labour cost.
Mandatory PZP periods come from the versioned legal ruleset. They remain locked and identical in both alternatives.
Total cost comprises role effort, non-labour cost, delay cost and explicitly monetised contract-design dimensions.
Informal process bypass is not monetised in model 2.3.
Every decision record exposes the assumptions, calculation anchors, evidence class, legal provenance and dimensions outside monetisation.

Result and range

Delta C = total cost of the formal sequential alternative - total cost of the adaptive compliant alternative

A positive value means that the formal sequential alternative has the higher declared cost. A negative value means that the adaptive compliant alternative has the higher declared cost. Low, central and high values form a scenario stress range, not a confidence interval. The outer range combines opposite endpoints of the two alternatives. Crossing zero does not by itself prove a sign reversal under shared assumptions.

Monetisation coverage

The native record separates role cost, non-labour cost, delay cost, competition transfer, contract amendments and TCO. Contract-amendment and TCO differentials remain zero in model 2.3, including edited scenarios. Informal bypass is reported as a non-monetised dimension.

Evidence and transfer limits

The evidence register distinguishes empirical anchors, official examples, retained scenario assumptions, internal illustrative allocations, practitioner observations and research hypotheses. Source relevance does not by itself justify numerical transfer.

Szucs provides the bounded empirical anchor for a competition-to-price channel in Hungarian procurement below the studied threshold. The 2, 6 and 9 per cent range is an explicit stress transfer, not an estimate for Poland.
Official material from UZP, the European Commission, OECD and California describes mechanisms such as market consultation, problem definition and modular contracting. It does not calibrate ProcuraCost cost effects.
Economic values, hourly rates, aggregate day totals, system-support multipliers and coordination or tool-cost profiles retained from model 2.2.2 remain labelled historical assumptions. For the five mechanism scenarios, step order, day allocation and role-hour allocation are illustrative model 2.3 inputs; external cases support constructs only, not duration, effort or cost.

Practitioner-material boundary

Procurement&Beyond episode 8 informs questions about process friction, internal ownership, requirements, TCO and responsible automation. It supports question design and hypothesis generation only. It does not calibrate thresholds, weights, ranges or expected implementation effects.

Scope and limitations

The legal ruleset covers the stated classic procurement contexts initiated in 2026 or 2027. Sectoral, defence and security contexts fail closed.
Procedure-family suitability is compared separately and without scoring. Special procedures still require a competent legal assessment of their statutory grounds.
The output is an auditable cost record for scenario analysis. It is neither legal advice nor statistical validation of an organisational outcome.

Selected references

The model register, deterministic replication artefacts and parameter documentation are maintained in the project repository.