Government proposals get weighed with a simple test: add up every cost and every benefit the project imposes on society, convert future dollars and pounds into present value, and check which option leaves the largest surplus. That is cost benefit analysis explained in one breath. The arithmetic is easy. The assumptions are not.
The method is formal doctrine, not habit. The White House Office of Management and Budget revised Circular A-94 in November 2023 to tell federal agencies how to judge public investments, and its stated goal is promoting social welfare through well-informed decisions. Britain keeps its own rulebook, the Treasury's Green Book, for the same purpose.
What Does Cost Benefit Analysis Actually Measure?
It measures net effects on society, not on the treasury alone. Per A-94, agencies run benefit-cost analysis on federal activities whose benefits and costs arrive spread over time. Analysts list the effects they can value, price them, discount future years back to today, and subtract. A positive total means the proposal passes. This connects to our earlier piece, Analysis Group: What Policy Analysts Actually Do All Day.
The rulebook also knows its limits. When an effect resists pricing, A-94 tells analysts to run a break-even test and ask whether the unpriced benefits could credibly justify the project anyway. That instruction is an admission on paper. The spreadsheet never captures everything, and its authors say so.
Why Do Discount Rates Change the Verdict?
A benefit promised in forty years counts for less than a benefit paid today, and the shrink rate is a policy choice. OMB's Circular A-4 estimates the social rate of time preference at 2.0 percent per year, built from a 1.7 percent real return plus 0.3 point added for inflation measurement.
Britain weighs the future differently. The Green Book sets its rate at 3.50 percent in real terms for years one through thirty, 3.00 percent for years 31 to 75, and 2.50 percent beyond that. The same road can pass in one country and fail in the other. The verdict moves because the clock is weighted, not because the road changed. A-94 expects sensitivity runs for exactly this reason: analysts recompute outcomes under alternative rates and show where the ranking flips.
Which Assumptions Do the Heavy Lifting?
The rate is only one loaded dial. Agencies also price prevented deaths, saved travel hours, cleaner air, and quieter streets, and each price comes from studies that reasonable people dispute. Move any of those inputs by a fraction and the ranking of options moves with it. The assumptions decide close cases long before the math runs.
A-94 adds a fiscal wrinkle. Because taxes distort choices, agencies may present a supplementary case that multiplies public spending by 1.25, reflecting an estimated 25 cents of added cost per dollar of revenue. The instruction is not decoration. Add a quarter to the cost side and a marginal project dies. One assumption quietly settled the verdict. For related coverage, see Twenty-Four Languages, One Law: What Multilingual Parliaments Actually Cost.
Where Does the Method Falter?
The method's defenders make a fair case. Imperfect numbers still beat raw lobbying, because the assumptions get published and can be attacked in public before the money moves. A written rate invites argument. A closed room does not.
The honest critique starts with A-4's own words: in practice, it is often difficult to quantify and express all of the important effects of a policy. Net totals hide who gains and who loses. The comparison baseline, meaning what happens without the project, is itself a modeling choice. And when a government grades its own proposal, the numbers are that government's position, not an independent audit.
Conclusion: A Debate in Spreadsheet Form
Cost benefit analysis is a formal argument, not a truth machine. It forces agencies to write down what a life, an hour, and a future year are worth, and it lets outsiders contest each figure. The totals deserve respect and suspicion in equal measure. Read the rate and the prices first. The conclusions follow from them.
