Offering · for organisations
Where a criterion enters a decision determines whether it can change the outcome.
An organisation decides to factor sustainability into its portfolio decisions. A year later the portfolio looks much as it would have looked anyway, and nobody can say precisely why.
The usual diagnosis is that the criterion was not weighted heavily enough. That diagnosis is often wrong, and acting on it costs another year.
What is measured, and how. Carbon footprint in CO₂e. Value per unit cost. Procedural burden on participants. Change the measure and every number downstream changes. No decision has changed yet.
How much each criterion counts relative to the others. This is where almost all organisational argument happens, and where governance concentrates its attention. In the case I studied, it is also where the least was decided.
Which options are permitted into the comparison at all. A threshold, a floor, a minimum standard, a rule about what may be proposed. Nothing here looks like a funding decision. Everything here constrains one.
In the prespecified simulation in my thesis, changing the weight on sustainability rarely changed which portfolio led. The eligibility rules had already decided which portfolios could compete.
Illustrative. These are made-up portfolios chosen to demonstrate the mechanism. They are not output from the simulation in my thesis, and nothing here is evidence for how often eligibility dominates weighting in practice. That question remains open, and the thesis says so.
The limit of that finding. It is conditional on the construction studied. Whether eligibility dominates weighting in general remains open, and the thesis says so. What does not depend on the construction is the result below.
A target set against your own performance cannot require improvement.
A sustainability limit set at a portfolio’s own measured performance will admit that portfolio unimproved whenever three conditions hold: equality is permitted; the same measurement both sets the limit and tests against it; and the reference value stays fixed, or is recomputed in the same state.
The result requires no data. It follows from how such a rule is written and holds whatever the calibration. That is why it applies well beyond pharmaceutical portfolios, to any organisation that benchmarks itself against its own past.
The limit admits the portfolio that set it.
Holds whenever equality is permitted, the same measurement both sets and tests the limit, and the reference value stays fixed or is recomputed in the same state.
Portfolio 1 of 32
Five funded programmes, each delivered centrally or decentrally. Two options across five programmes give thirty-two portfolios, and because thirty-two can be enumerated exhaustively the framework evaluates every one. The portfolio, not the programme, is the unit of decision, because the budget cap and the sustainability ceilings apply to the combined slate. The five programmes are archetypes I constructed to span the declared decision space; they are drawn from no observed pipeline.