This post was initially written in French https://freakonometrics.hypotheses.org/87034
The idea that the economy is a zero sum game is remarkably persistent. It comes up almost every time we talk about international trade, taxation, insurance, or redistribution. If someone gains more, someone else must necessarily lose. This intuition feels natural because it matches many everyday situations. When a price goes up, the buyer feels like they are losing what the seller is gaining. When an insurance premium is paid, the policyholder feels like they are funding other people’s claims. This intuition is also deeply anchored in our relationship with accounting. We learn very early that accounts have to balance. For a household, a firm, or a state, what is spent must match what is received. Spending more than revenue is perceived as an anomaly or a fault. Reasoning in terms of balanced accounts naturally leads us to reason in terms of flows that cancel out. Every expense mirrors a receipt. Seen from that angle, the economy looks like a vast system of transfers where the gains of some necessarily correspond to the losses of others.
Accounting reinforces this zero sum idea because it records monetary flows, not the indirect effects of decisions. It is indispensable for financial coherence, but it says nothing about value creation, the reduction of uncertainty, or risks that are avoided. An investment in prevention appears as a cost today, while the losses it prevents will never show up in the accounts. Likewise, a well designed insurance system can stabilize individual and collective trajectories without ever producing a visible accounting profit. This is exactly where the zero sum intuition becomes misleading. The economy is not reducible to accounting, even if it must live with it. Externalities, prevention, and risk management introduce mechanisms through which individual decisions change the size of the pie itself. Insurance is a privileged setting to observe this gap between accounting logic and economic logic. It looks like a zero sum game when we focus on the flows, but it can become a positive sum or a negative sum game when we focus on the behaviors it induces and the risks it avoids or amplifies.
Economics and the zero sum game
One of the oldest and most robust messages of economics is precisely that the zero sum intuition is often wrong. In An Inquiry into the Nature and Causes of the Wealth of Nations Recherches sur la nature et les causes de la richesse des nations in French, Adam Smith explains that the division of labor and voluntary exchange can increase total production. In On The Principles of Political Economy, and Taxation, David Ricardo shows that even when a country is more productive in every activity, trade can still be beneficial thanks to comparative advantage. Exchange is not the sharing of a fixed stock. It can create value. This idea has become a pillar of modern economic teaching. Gregory Mankiw puts it at the heart of Principles of Economics, reminding us that trade allows everyone to specialize in what they do best. Yet, as Robert Armstrong notes in The essence of economics, published in the Financial Times earlier this month, zero sum reasoning keeps coming back, especially when value creation mechanisms are indirect, diffuse, or hard to observe. Part of the explanation lies in how naturally we reason in accounting terms. We learn very early that accounts must balance. At the scale of a household, a firm, or a state, flows seem to cancel out. An expense looks like a loss, a receipt looks like a gain. Reasoning in terms of balanced accounts almost mechanically leads to zero sum reasoning.
As Tim Harford reminded us at the start of the summer in The zero-sum mindset is no mystery, this way of seeing is not only a reasoning mistake. It is often the product of real experiences that feel zero sum. When growth is weak, when opportunities are scarce, when resources appear limited, the world does indeed look like a game of winners and losers. Tim Harford notes that zero sum reasoning is more frequent among individuals and societies that have experienced little growth, tight labor markets, or intense competition for resources perceived as scarce. The central question then becomes the entity over which the sum is computed. At the scale of an indebted household, inflows and outflows do not line up, and the feeling of loss is very real. At the scale of a firm that invests, costs are immediate while benefits are uncertain and delayed. At these levels, the sum can be negative. But at the scale of an ecosystem, the very same flows can correspond to investments, avoided risks, or increased capacities that benefit other agents or society as a whole. It is precisely in this gap between the accounting logic of individual agents and the logic of the system that externalities play a central role. When an individual decision has effects on others that do not pass through prices, the zero sum intuition becomes especially misleading. Benefits and costs no longer coincide with the accounting boundaries of those who make decisions.
Insurance can be an example of this gap. From an accounting perspective, it looks like a redistribution game. Premiums finance claims, and taken in isolation the books balance. But at the scale of the ecosystem, insurance profoundly changes how risk is produced, distributed, and sometimes avoided. It influences behaviors, prevention, and the stability of individual and collective trajectories. Understanding insurance therefore requires moving beyond the logic of isolated accounts and reasoning in system terms. This is why it is such a fertile laboratory for thinking about value creation, externalities, and the common good.
Value creation and positive externalities
Or why the pie can genuinely grow. The economy stops being a zero sum game as soon as individual actions produce effects that go beyond the strict frame of the transaction. Arthur Pigou formalized this idea in The Economics of Welfare by introducing the notion of externality. A positive externality exists when an action provides benefits to others without compensation. A negative externality appears when costs are imposed on others without compensation. Classic examples of positive externalities include vaccination, which protects the individual and the community, education, which increases individual productivity and overall growth, and scientific research, which generates knowledge that diffuses far beyond those who produce it. In all these cases, the sum of social benefits exceeds the sum of private benefits.
Externalities also help explain why some policies can have the opposite effect and reduce collective welfare. A concrete and recent example is the introduction of tariffs. Presented as a way to protect certain domestic producers, tariffs are a negative sum game in the economic sense. They raise prices for consumers, increase costs for firms that use imported inputs, and create distortions in production. Gains for protected sectors do not offset losses elsewhere. As Mary Amiti, Stephen Redding, and David Weinstein show in The Impact of the 2018 Tariffs on Prices and Welfare, value is literally destroyed, in the form of deadweight losses and additional coordination costs.
Arthur Pigou starts from a simple reasoning. When an activity imposes a cost on others without compensation, for instance a factory that pollutes a river, the market does not lead to an efficient allocation. The state should intervene through a tax or regulation to force the polluter to account for the damage. Ronald Coase, in The Problem of Social Cost, does not say Pigou is wrong. He shows that the problem is not only pollution, but the absence of clear rules about who has the right to do what, and the cost of coordination between parties. If property rights are well defined and bargaining costs are low, actors can sometimes reach an agreement without external intervention. The classic example is a factory and a neighbor. If the factory has the right to pollute, the neighbor can pay compensation for a reduction in pollution. If the neighbor has the right to clean air, the factory can compensate the neighbor to keep producing. In both cases, if negotiation is feasible and cheap, an efficient solution can be reached. The externality is internalized through private agreement rather than a tax.
Coase’s central point is that externalities are coordination problems and their treatment depends heavily on transaction costs. When these costs are high, as in climate or public health, private solutions become unrealistic and collective institutions are needed. Elinor Ostrom showed in Governing the Commons that such institutions are not limited to a market versus state opposition. Forms of collective governance can emerge to manage common resources and internalize externalities. William Nordhaus illustrates this in The Climate Casino, showing that fighting climate change requires large scale coordination mechanisms combining economic incentives, regulation, and insurance. These works show that value creation is not limited to market exchange. It depends on the ability of societies to organize the management of externalities. This is exactly where insurance becomes interesting.
Insurance as risk sharing
At first glance, insurance seems to confirm the zero sum intuition. Premiums collected pay claims. What some pay, others receive. On average, the system looks closed. This is the definition I give of insurance in my classes.
the contribution of the many to the misfortune of the few.
This reading is deeply incomplete. In Essays in the Theory of Risk Bearing, Kenneth Arrow shows that the value of insurance does not lie in changing the expected loss, but in reducing uncertainty. Two situations can have the same average loss but very different risks. For risk averse individuals, variance has a welfare cost, which a simple accounting equation does not capture. Jan Mossin, in Aspects of Rational Insurance Purchasing, formalizes this intuition by showing that demand for insurance depends on risk aversion and the structure of uncertainty. Pooling transforms highly volatile individual losses into more predictable contributions. The value created is real, but it is invisible in the accounts.
Harry Markowitz showed that this reasoning is exactly that of portfolio diversification, developed in Portfolio Selection. Reducing risk without reducing expected return is a source of value. Insurance applies this logic to non financial risks. However, this value creation rests on key assumptions. Risks must be sufficiently independent. When losses are correlated, as in natural disasters or pandemics, aggregate variance does not disappear. Chi, Zhou, and Zhuang show in Variance insurance contracts that modern insurance is constrained by aggregate volatility. Solvency becomes a scarce resource, and its management directly affects premiums and coverage. So even before we talk about behavior, insurance is already not a simple transfer game. It changes the distribution of risk and creates value by reducing uncertainty. But even this dimension is still incomplete.
When insurance creates collective value
Insurance does not only spread existing losses. It also changes behavior, and therefore changes the production of risk itself. This is where positive externalities appear most clearly. In their foundational article Market Insurance, Self Insurance and Self Protection, Isaac Ehrlich and Gary Becker distinguish three mechanisms. Market insurance transfers risk. Self insurance consists in saving to face losses. Self protection aims to reduce the probability or severity of losses. These three dimensions interact. Contrary to a common idea, insurance does not necessarily reduce prevention.
Insurance contracts include many incentive mechanisms. Deductibles, bonus malus schemes, safety audits, and standards imposed by insurers can encourage more cautious behavior. When a driver adopts safer driving, they reduce not only their own risk but also the risk of other road users. When a building is better protected against fire, neighboring buildings also benefit from reduced risk. Annette Hofmann shows in Internalizing Externalities of Loss Prevention through Insurance that insurance can be analyzed as an instrument to internalize prevention externalities. By adjusting premiums and coverage conditions, insurers align individual incentives with the collective interest.
Howard Kunreuther and Erwann Michel Kerjan develop this idea in At War with the Weather. They show that premiums that better reflect risk, combined with investments in prevention, can reduce future damages from natural disasters. In that case, insurance does not just redistribute losses. It helps prevent losses that had no good reason to occur. In health, Mark Pauly explains in Medical Care at Public Expense that insurance can encourage preventive behavior when it facilitates access to early care. Spending today then avoids much higher costs tomorrow. Insurance becomes a mechanism of dynamic value creation.
When insurance destroys value
The same mechanisms can produce the opposite effects. Insurance can also generate negative externalities and reduce collective welfare. Moral hazard is the best known mechanism. When an individual is protected from the financial consequences of their actions, they may adopt riskier behavior. Mark Pauly showed that this phenomenon is particularly important in health insurance, where coverage can lead to overconsumption of care with low marginal value.
Adverse selection is another central problem. In Equilibrium in Competitive Insurance Markets, Rothschild and Stiglitz show that when individuals know their risk better than insurers do, the market can segment inefficiently. Contracts attractive to good risks disappear, and overall coverage deteriorates. Liran Einav and Amy Finkelstein synthesize these results in Selection in Insurance Markets Theory and Empirics in Pictures. They show that selection and moral hazard are pervasive, but vary in magnitude across contexts. They also note that excessive segmentation can weaken solidarity and reduce the social acceptability of insurance.
Deborah Stone adopts an institutional perspective in The Limits of Professional Power. She shows that actuarial pricing, pushed to the extreme, can turn insurance into a mechanism of social sorting. By trying to make everyone pay strictly according to their risk, we can exclude the most vulnerable and make some risks uninsurable. In these situations, insurance does not just redistribute losses. It changes behavior and social structures in a way that can be collectively costly.
Insurance, fairness, and the common good
These tensions show that insurance is a social institution as much as an economic mechanism. It rests on explicit tradeoffs between efficiency, fairness, and solidarity. In Creating Capabilities, Martha Nussbaum proposes analyzing these questions through people’s real capabilities. Risk and vulnerability limit the possibility of living the life one has reason to value. Insurance can be seen as an infrastructure of freedom, but this function requires some pooling of risks. Jean Tirole insists in Économie du bien commun that economic institutions must be evaluated against the common good. Efficiency is not enough. Social legitimacy and political sustainability matter.
More recent work, such as Better Together How Externalities of Size Complicate Notions of Solidarity and Actuarial Fairness by Donahue and Barocas, shows that the increasing segmentation of risks raises new problems. Size externalities and complex correlations make pricing justice harder to define. Insurance then reveals a central tension between individualization and solidarity.
Conclusion
Insurance illustrates quite simply why the economy is not a zero sum game. In accounting terms, flows may look balanced. But economically and socially, everything depends on the behaviors insurance induces, the risks it avoids, and the institutions that frame it. When it reduces uncertainty, encourages prevention, and internalizes positive externalities, insurance creates collective value. When it fosters moral hazard, excessive selection, or social fragmentation, it can instead destroy value. It is neither intrinsically virtuous nor intrinsically perverse. Understanding insurance means understanding that the relevant question is not who pays for whom, but which rules make it possible to turn a transfer mechanism into an institution that creates welfare. This distinction is precisely what helps move beyond the misleading zero sum intuition and think of the economy as a set of collective choices in the face of risk.
OpenEdition suggests that you cite this post as follows:
Arthur Charpentier (December 29, 2025). Insurance, a Zero Sum Game? Freakonometrics. Retrieved January 13, 2026 from https://doi.org/10.58079/15fi1