Why prevention
The village keeps books.
Everyone in the impact economy knows the parable of the villagers pulling drowning strangers from the river until someone walks upstream. It is a good parable. It is missing its most important page: it never mentions that the village keeps books.
The unread ledger
Somewhere in that village is a ledger, recording the cost of the boats, the watchers on the bank, the healers, the burials, the rebuilding after every season of bodies. Read it and you find the strangest entry of all: vast sums spent hauling consequences out of the water, and comparatively little spent walking upstream. The parable presents prevention as a moral awakening. The ledger reveals it as an accounting correction, which is, for institutions, a far more actionable thing to be.
Prevention compounds; reaction repeats.
Budgets are like that. An institution’s budget is the truest autobiography it will ever write. Mission statements record what an institution wishes to believe; allocations record what it actually believes, committed annually, defended politically, signed in ink. Read as autobiography, many institutional budgets reveal the same tilt: far more money allocated to absorbing consequences than to reducing their likelihood. In the book’s estimate, institutions often direct just 10 to 20 percent of their security-related spending to prevention, depending on the perimeter and horizon chosen. The band is not yet a benchmark. It is a signal.
The pattern runs from the top of the world’s accounts to the bottom. Global military expenditure reached approximately $2.9 trillion in 2025, the eleventh consecutive year of growth, pushing the world’s military burden to its highest share of global GDP since 2009. Beneath that reactive summit, every city, company, university, and foundation maintains its own smaller fortress budget, mostly unread.
Here security is used in the broad sense: the systems that protect people, places, institutions, and enterprises from social, ecological, economic, and physical harm. That broader definition matters because the prevention–reaction split is usually hidden across organizational silos. Economists have studied the economics of prevention for decades, particularly in public health. What has been missing is a cross-domain discipline that asks the same allocation questions of security, health, climate, ecology, finance, and enterprise: how much do we spend before harm versus after it? What value follows when that allocation changes? Who captures it? Call the discipline Prevention Economics: the study and practice of how institutions allocate resources between reacting to harm and preventing it, and of the value created when that allocation begins to move.
The two numbers
You cannot steward what you refuse to count; you cannot move what you have never measured.
Prevention Economics begins with a measurement so simple it is almost embarrassing that institutions do not already publish it. The Prevention Ratio™ is the share of security-related spending directed to prevention rather than reaction: preventive spending divided by the sum of both. Expressed in cents, it asks a question anyone can hold. Of every security dollar, how many cents go upstream? The ratio is not a claim of perfect classification. It is a declared estimate, bounded by an institution’s chosen perimeter and time horizon. Its power is not false precision but legibility: it converts a diffuse institutional instinct, “we are too reactive,” into a figure a board can see, interrogate, track, and move.
The second number describes what movement is worth. The Prevention Dividend™ is the economic value generated when spending shifts from reaction to prevention. Unlike the Ratio, it is not one standardized number. It is a family of return measures, benefit–cost ratios, annual social rates of return, and averted-loss estimates among them, that answer different questions and must travel with their class, source, horizon, and limits. Nor is the claim that prevention automatically pays for itself. It does not. Poorly designed or badly targeted prevention can waste money as surely as poor reaction can. The stronger claim is that well-designed preventive investments have repeatedly generated substantial economic value across independent domains.
The evidence, by class
Five domains, four kinds of number.
Each figure is labeled with the kind of number it is and where it comes from.
Benefit–cost ratio
$4–$7 per $1
Natural-hazard mitigation. Twenty-three years of U.S. federal mitigation grants returned about $6 for every $1 spent, and about $7 for riverine flooding. Building-code improvements returned about $4.
Benefit–cost ratio
$7–$30 per $1
Ecosystem restoration. Restoring degraded land has been estimated to generate $7 to $30 in economic benefits per dollar invested.
Ding et al., World Resources Institute, 2018 (opens in a new tab); corroborated by OECD, 2019 (opens in a new tab)
Annual rate of return
7–13%
Early childhood. High-quality early-childhood programs have produced estimated annual social returns of roughly 7 to 13 percent, depending on the program. A rate of return, never a dollar multiple.
Heckman et al., 2010 (opens in a new tab); García et al., 2020 (opens in a new tab)
Averted-loss estimate
$1,102 per $1
Pandemic preparedness. Modeling of a U.S. COVID-like scenario estimated that each dollar invested in preparedness could avert $1,102 in expected GDP loss. An averted-loss estimate, not a benefit–cost ratio.
Jameel Institute, Imperial College London (opens in a new tab)
Modeled cost-effectiveness ratio
$16 per $1
Violent-conflict prevention. A scenario that more than doubles peacebuilding funding in the 31 most fragile and conflict-affected countries for a decade estimates that each $1 would reduce the cost of conflict by $16 over the long run. A projection of what prevention could save if it works as modeled, not a measured return.
These figures are not interchangeable. Their convergence is directional: across very different systems, moving resources upstream can create value that conventional budgets systematically fail to see.
Real security
Five pillars, not one wall.
Real security is not a fortress. It is a web, built from the interwoven systems that make collective life possible: the health of ecosystems that buffer disasters, the strength of public institutions that maintain trust, the equity of economic systems that prevent desperation, the resilience of communities that care for their own. Weaken any one pillar and the whole becomes brittle; strengthen all five and threats become less likely to emerge and easier to absorb when they do.
Ecological stability
Healthy ecosystems are the original security infrastructure. They regulate climate, purify water, regenerate soil, buffer storms, and support the biodiversity on which everything else depends.
Public health capacity
Pandemics, antimicrobial resistance, mental-health crises, chronic disease: these are security threats that kill more people and cost more than most conventional conflicts, and a system that can prevent them is a system that can absorb shocks.
Economic equity
Inequality is a threat multiplier. Economic security is not everyone equally wealthy; it is everyone with access to dignified work, basic needs, and a path to stability.
Democratic infrastructure
Democracy is not a luxury; it is a security technology. Where people trust that their voices matter and that conflicts can be resolved through dialogue, stability deepens.
Community cohesion
The most invisible pillar, and the one that activates all the others: the relationships of trust and reciprocity that let people organize mutual aid, protect the vulnerable, and hold institutions to account.
Seven principles for a young field
If Prevention Economics is to become a commons rather than a brand, it needs principles anyone can hold its practitioners to.
- Read the ledger first. Every institution’s true security doctrine is already written, in its allocation. You cannot steward what you refuse to count; you cannot move what you have never measured.
- Not all returns are the same kind of number. A benefit–cost ratio, an annual rate of return, and an averted-loss estimate answer different questions. The Prevention Dividend is a discipline of classification before it is a claim of return.
- Prevention compounds; reaction repeats. Reactive spending purchases the same year over and over. Preventive spending can purchase different futures: avoided losses that stack, capacities that persist, trust that accrues.
- The wrong pocket is often the binding barrier. The actor who pays for prevention is often not the actor who captures the savings. Underinvestment in prevention is therefore often a routing failure, not a knowledge failure.
- The optimum is not one. A ratio of 1.0 would be as pathological as zero. Prevention Economics seeks the right ratio, not the maximal one, and admits plainly that the theory of the optimum remains unwritten.
- Open the standard; verify the claims. The definitions belong to everyone: a metric no one else can compute is a service, not a standard. What must be guarded is the integrity of its use: declared perimeters, declared horizons, transparent classification, verified rather than self-reported claims.
- Every reactive line item is a design brief. Read backward, the ledger is a map of things waiting to exist.
Three practitioners, one ledger
An institution’s budget is the truest autobiography it will ever write.
A field earns its keep by serving people with different jobs. Institutions read the ledger. Funders steer across many ledgers. Founders originate from them. For a city manager, the ratio provides what no strategic plan does: a visible baseline, a declared perimeter, and a direction of travel that can survive the next election. The debate shifts from whether to prevent, which everyone favors the way everyone favors good weather, to the only question a budget can answer: what should the next five cents buy? For a foundation, the ratio of its own grantmaking may reveal something it did not plan to know: that for all its upstream language, its money is repeatedly pulled toward visible crises, where relief outruns resilience and treatment outruns prevention. The field’s hardest puzzle follows: prevention’s returns often land in the wrong pocket, and repairing that misalignment is the design space of prevention finance. For a founder, a reactive line item is pain already quantified, priced, renewed annually and politically defended: a commitment, not an opinion. Whose reactive spending does this venture convert into prevention? Ask it, and the ledger becomes a map.
The open problems
A young field should be judged by the problems it admits. Four stand open.
Classification is perimeter- and horizon-relative: insurance finances recovery yet prices risk; maintenance is prevention relative to failure and operations relative to design. Ratios are comparable only under declared perimeters, declared horizons, and transparent rules, and that definitional work belongs in peer-reviewed literature and open standards. There is no theory yet of the optimal ratio, which will vary with hazard profile, insurance markets, mandate, and existing capacity. The dividend’s epistemics are hard: averages are not marginals, avoided costs rest on modeled counterfactuals, and published ratios skew toward programs that worked. Honest practice speaks in ranges with stated confidence, distinguishes realized from modeled returns, and never turns a point estimate into a promise. And Goodhart waits at the door: the day the ratio becomes a target, institutions will reclassify spending to look preventive. Fields are constituted by benchmarks, and benchmarks survive by verification.
These are not footnotes. They are the research agenda, and they are open.
Institutions that move early stand to capture the Prevention Dividend while others continue paying the escalating costs of reaction.
A first conversation is a diagnostic, and it costs nothing.