When Numbers Start Making Decisions · Season Four, “Things That Cannot Be Priced Directly” · Article 11
1. What is A$1 million in 30 years worth today?
Suppose a policy will prevent a loss of A$1 million in 30 years. At a real annual discount rate of 7 per cent, the present value is about A$131,000. At 3 per cent it is about A$412,000; at 10 per cent, about A$57,000. The future event is unchanged, but today’s cost-benefit conclusion can change completely.
The Australian Office of Impact Analysis requires a central real discount rate of 7 per cent for net present value in ordinary cost-benefit analysis, with sensitivity analysis at 3 and 10 per cent. For very long environmental timeframes, it recommends declining rates: 7 per cent for years 1–30, 5.4 per cent for years 31–75, then lower bands. OIA: Environmental valuation
Discounting is often explained by saying that future money is worth less than money today. When the future benefit is avoiding death, ecological collapse or climate harm, that sounds like future people matter less. Those propositions are not identical.
The question is what a discount rate legitimately converts and where it exceeds its authority.
2. Discounting places different years at one decision point
Money available now can be invested and produce returns. Spending it on one policy means forgoing other projects. People may prefer benefits sooner and costs later, while future forecasts carry uncertainty. Discounting converts future costs and benefits into a base-year present value so options with different timing can be compared.
The standard formula divides a future amount by one plus the discount rate raised to the number of years. Compounding means the effect becomes dramatic over long periods. A small change in the rate can dominate assumptions about distant harms.
For monetary flows, this is necessary. Adding A$1 million today directly to A$1 million in 2056 ignores opportunity cost and time. The dispute concerns which rate reflects the social opportunity cost and whether the same conversion should apply to every consequence.
Inflation must be kept separate. A real discount rate is applied to amounts expressed in constant prices; a nominal rate belongs with amounts that include expected inflation. Mixing a real rate with nominal cash flows can bias the result before the ethical debate even begins. Analysts should state the price year and show that costs, benefits and rate use the same basis.
OIA guidance requires the range precisely because there is no single observable “true” social rate. OIA: Cost Benefit Analysis
3. Seven per cent is not a scientific boundary in future value
The 7 per cent central rate is a policy convention for consistent analysis, informed by opportunity-cost reasoning. It is not a finding that every value loses seven per cent of moral importance each year.
Other frameworks use different rates depending on health, infrastructure, climate and public finance. The correct rate can depend on whether funds displace private investment, public consumption or another project, and on risk characteristics.
Consistency helps prevent analysts from choosing a favourable rate for each proposal. Sensitivity analysis reveals when the recommendation depends on the convention. If the net result changes sign between 3, 7 and 10 per cent, the decision-maker should not describe the project as plainly efficient.
The rate is an input to judgement. Hiding it inside a spreadsheet turns an ethical and economic choice into invisible arithmetic.
4. Future money and future people
Discounting a monetary amount does not necessarily discount the moral standing of the person affected. If society invests resources today, future people may inherit more productive capacity. Present value asks how much current resource is equivalent to a future monetary flow under stated assumptions.
But intergenerational projects complicate that defence. Future people cannot bargain today, and environmental losses may be irreversible. Economic growth may make future generations wealthier on average while climate damage or extinction leaves them with fewer options. Distribution within generations also matters.
Some consequences should be represented in physical units and ethical constraints alongside discounted money. Tonnes of emissions, extinction risk, lives affected and cultural loss should not disappear when their monetary present value becomes small.
Declining long-term rates recognise uncertainty about the future discount rate and reduce the extreme shrinkage caused by applying one high rate for centuries. They do not solve every intergenerational question, but make distant effects harder to erase.
5. The discount rate shapes project design
A high rate favours early benefits and delayed costs. It can make preventive maintenance, resilience and ecosystem restoration look unattractive because their benefits arrive later. It can reward a design with low initial cost and high future remediation.
A lower rate gives greater weight to long-lived benefits and distant liabilities. It can make climate action, durable infrastructure and early prevention more favourable. Neither result proves the project is good. It shows that timing is part of value.
Analysts should report undiscounted time profiles as well as present totals. Decision-makers need to see when costs and benefits occur, who experiences them and whether the project creates a tail of irreversible harm.
6. Uncertainty should not all be hidden in a higher rate
Future benefits can be uncertain, but increasing the discount rate is a blunt response. It penalises all future outcomes exponentially and can double-count risk if expected values already incorporate probability.
Different uncertainties need different treatment: probability distributions, scenarios, sensitivity analysis, real-options analysis and adaptive policy. Climate tipping points and irreversible loss may justify precaution rather than a higher rate that makes them numerically negligible.
The OIA’s environmental guidance recommends balanced presentation, important caveats and adaptive strategies where information may change. A confident single net present value is inappropriate when the underlying ecological model is deeply uncertain.
7. Who can challenge the rate?
Impact analysis should publish the base year, real or nominal basis, time horizon, rate path and sensitivity results. Stakeholders should be able to identify which future effects drive the result and reproduce the calculation.
Review should distinguish arithmetic error from disagreement about assumptions and ethical constraints. A proposal may pass at every rate but still violate a legal environmental limit. Another may fail at 7 per cent but deliver a non-monetised strategic obligation.
The responsible minister or decision-maker must explain why the chosen analysis fits the policy. “The spreadsheet discounted it” is not an account of responsibility.
8. The future is not a smaller version of the present
Future conditions, technologies, populations and ecological thresholds can differ qualitatively. Present-value calculation treats time through one mathematical relation, which is useful for comparable resource flows. It cannot contain every transformation across generations.
The rate should discount resources where appropriate, while rights, irreversibility and minimum duties remain constraints outside compensation. A large current benefit should not automatically purchase an irreversible future loss merely because the latter’s present value is small.
9. A practical test for long-term discounting
Ask:
- Are amounts real or nominal, and is the rate consistent?
- What opportunity cost does the rate represent?
- Does 3–7–10 per cent sensitivity change the decision?
- Is a declining long-term rate required?
- Are physical and distributional outcomes reported by year?
- Is uncertainty modelled separately rather than hidden in the rate?
- Are irreversible harms subject to non-monetary constraints?
- Can future obligations be adapted and reviewed?
Conclusion: discount resources, not irreversible responsibility
Discounting is essential for comparing resource flows across time. A 7 per cent central rate with 3 and 10 per cent sensitivity creates a consistent starting structure for Australian impact analysis.
My judgement is that no rate should be allowed to make distant people or irreversible ecological duties disappear. Long-term analysis needs declining rates where applicable, transparent timing, separate uncertainty and constraints that money cannot offset.
The future amount may have a smaller present financial value. The future person does not have a smaller claim to institutional responsibility.
Discover more from Geoffrey Chen
Subscribe to get the latest posts sent to your email.