ProcuraCost methodology notes

Published materials

1. 01

   Methodology

   ProcuraCost 2.3.0: what does the cost model compare?

   Procurement workflow design, competition and contract design are separate dimensions. The model does not replace them with a composite index or technology label.

   Focus: Calculation contract

   Procurement practice implication: Before calculation, document the activity maps, competition exposure and contract design separately.

2. 02

   Competition and contractor selection

   The Szucs study: discretion, competition and price

   In the studied Hungarian market, the structural model indicates approximately 6 per cent higher prices and selection of contractors with 28 per cent lower productivity as measured in the study. The study covers contracts below an approximately HUF 25 million threshold. Model 2.3 uses only a bounded transfer of the price channel as an explicit stress range.

   Focus: Source review

   Procurement practice implication: Measure competition exposure independently of workflow design. Transfer beyond the studied population requires an explicit assumption.

   Szucs (2024), primary publication ↗
3. 03

   Contract design and formal amendments

   The Beuve study: contract design and formal amendments

   Table 4 of the published version implies 0.077-0.098 additional formal amendments per contract-year for a simultaneous one-standard-deviation increase in each of seven rigidity categories in French car-park contracts; the 7.7-10.5 per cent prose is a working-paper leftover. It is a frequency, not an event probability or an effect of workflow design. In model 2.3 the differential is anchored to the verified post-print and monetised only under a declared contract-rigidity difference.

   Focus: Source review

   Procurement practice implication: Record clause design independently of activity count, role effort and procurement duration.

   Beuve, Moszoro and Spiller (2023), primary publication ↗
4. 04

   Result range

   How should the scenario range be interpreted?

   The low, central and high range tests declared assumptions, not statistical confidence. The outer range crossing zero does not prove a sign reversal under shared assumptions.

   Focus: Interpretation boundary

   Procurement practice implication: Interpret the central result together with cost drivers, the outer range and the local assumptions register.