Game Theory · GT-17
Everyone benefits whether or not they pay — so the individually rational move is to let someone else pay, and everyone reasons the same way.
A public good is non-excludable (you can't stop anyone from benefiting) and non-rival (one person's use doesn't reduce another's). Because individual contribution is costly but benefits are shared regardless, each rational individual is tempted to free-ride — leading to systematic under-provision relative to what the group would collectively prefer.
Formal treatment traces to Paul Samuelson's 1954 paper 'The Pure Theory of Public Expenditure,' building on earlier public-finance discussions of non-excludable goods.
The Mechanism
Contribution collapses as group size grows
As group size grows, individual contribution reliably falls — each person's share of both the benefit and the credit shrinks, while the temptation to let others cover the cost grows, a pattern replicated across decades of public-goods lab experiments.
01 · NON-EXCLUDABILITY IS THE ROOT CAUSE
You get the benefit whether you pay or not
Unlike a private good, nobody can be effectively blocked from enjoying a public good (clean air, national defense, open-source software) once it exists — which removes the market's normal enforcement mechanism (pay or don't get it) that would otherwise price contribution correctly.
02 · THE PROBLEM SCALES WITH GROUP SIZE, NOT JUST INCENTIVE STRENGTH
Small groups often self-organize fine
Mancur Olson's influential 1965 analysis (The Logic of Collective Action) showed the free-rider problem is far worse in large groups than small ones — in a small group, each member's contribution is a large, visible share of the total, and social pressure and reciprocity can sustain cooperation. In large groups, individual contribution becomes nearly invisible, and free-riding becomes close to costless.
03 · SOLUTIONS TARGET THE INCENTIVE, NOT JUST THE APPEAL
Moral exhortation alone rarely fixes it
Effective real-world fixes — mandatory taxation for public goods, membership dues with excludable perks, matching-donation campaigns, or Elinor Ostrom's documented community-governance mechanisms (1990 Nobel-winning work) — all work by changing the actual payoff structure of contributing, not simply by asking people to be more generous.
Where It Fails / Inversion
Where it fails / inversion
Not every shared-benefit situation triggers free-riding as strongly as theory predicts — Elinor Ostrom's field research documented numerous real communities successfully self-governing shared resources through locally-evolved monitoring and reputation systems, without top-down enforcement. Assuming free-riding is inevitable can lead to over-engineered, expensive top-down solutions where a lighter community mechanism would have worked.
How To Use It
Worked example · open-source software funding
Widely-used open-source infrastructure is a textbook public good — any company can use it freely, and no single user's non-payment excludes them from the benefit. This predictably under-funds critical infrastructure (documented repeatedly in projects like OpenSSL before major post-Heartbleed funding initiatives) until organizations like the Linux Foundation or corporate sponsorship consortia change the payoff structure — creating visible reputational and access benefits tied specifically to contribution, converting a pure public good into something closer to a club good.
How to use it
Before assuming voluntary goodwill will fund or sustain a shared resource your group depends on, calculate honestly whether it's genuinely a public good (non-excludable, non-rival) at meaningful scale — if so, design an actual incentive mechanism (dues, matching funds, visible recognition tied to contribution) rather than relying on appeals to fairness, which reliably underperform at scale.
See Also