Framework
Framework / multiplicative account
A fuller account of prosperity.
GDR treats ecological, social, and capital conditions as interdependent stocks and capacities that change through pressure, investment, depletion, recovery, and time.
Accounting grammar
Stocks, flows, pressures, additions, losses, and recovery.
Opening condition
What is true at the start of the accounting period.
Pressures
Forces that may degrade, constrain, or redistribute capacity.
Additions
Investment, restoration, capability-building, or other strengthening activity.
Losses
Depletion, damage, exclusion, extraction, or erosion.
Closing condition
What is true at the end of the period.
Direction
Whether the system is strengthening, weakening, stable, or unresolved.
System boundary
The same accounting questions can be asked at different scales.
Nation
A national account with explicit subnational and cross-border dependencies.
Bioregion
A system defined by ecological and human relationships rather than only political borders.
Organization
An institution examined through the capacities it depends on, affects, and can govern.
Non-compensatory aggregation
A gain in one pillar cannot automatically cancel a critical loss in another.
Most composite indices aggregate their components by weighted average, which assumes full substitutability: a deficit in one dimension can be compensated by a surplus in another. GDR combines its three pillars multiplicatively instead, so a severe weakness in any one of them drags the whole account down rather than being masked by strength in the other two. Within each pillar GDR uses the geometric mean, following the UNDP's 2010 revision of the Human Development Index, which was made precisely because the geometric mean does not allow perfect substitutability across dimensions. The lineage stops there: GDR's cross-pillar step is a raw product rather than a geometric mean, which is a stronger condition than the HDI imposes, and the two steps differ in kind: the within-pillar mean is partially compensatory, the cross-pillar product is strictly non-compensatory.
The claim, stated so it could be wrong
Pillar deficits widen the range of outcomes rather than lowering today's output.
The usual version of non-compensability, that no strength in one dimension can compensate for severe weakness in another, is either trivially true or unfalsifiable, depending on how long you are willing to wait. The claim GDR makes is narrower and testable: pillar deficits raise the variance and the left-tail severity of future output over multi-decade horizons, rather than reducing current output. A nation can run a severe deficit in one pillar and post strong output for a long time. What it cannot do, on this claim, is hold that position without accumulating exposure to a class of outcome a level-based measure does not price. The standing of the three parts of that claim is not the same, and is recorded here rather than levelled up.
Variance: substantial support
The resource-curse literature moved from a level effect to a volatility channel, and the volatility result survived the challenge the level result did not. Terms-of-trade volatility is associated with materially lower long-run per-capita growth, operating mainly through reduced capital accumulation.
Tail severity: mechanism, not estimate
The tipping-point and capital-withdrawal channels supply a mechanism by which a deficit produces a discontinuity rather than a proportional decline. No study located for the paper estimates pillar deficits against the third moment of the growth distribution. Closing that gap is the most valuable single empirical contribution the framework could make.
Substitutability: domain-specific
The strongest available evidence is empirical rather than axiomatic, and it supports low-to-moderate substitutability rather than none. Nature's provisioning services have permitted substitution; its regulating and maintenance services are complementary to one another. That is a weaker claim than a blanket one, and a considerably more defensible one.
Next
Explore the three pillars.
Each pillar carries its own evidence questions, limits, and relationships to the other two.