Free riding
Enjoying a shared benefit without paying your share.
What it means
Free riding is benefiting from a collective or public good while contributing little or nothing, relying on others to bear the cost. The mechanism is a misalignment between individual and collective incentives: because public goods are non-excludable, each person enjoys the benefit whether or not they pay, so the rational move is to let others contribute, even though all are worse off if everyone reasons the same way. It sits at the core of collective-action problems — Olson's argument that, absent selective incentives or coercion, large voluntary groups struggle to provide the goods their members want, though his further claim that larger groups always fare worse is not consistently borne out in experiments. Behavioral research finds that real people free-ride less than self-interest predicts, contributing substantially in public-goods experiments at first, though contributions typically decay over repeated rounds unless sustained by communication, reputation, or punishment. It relates to social loafing and diffusion of responsibility, and it shapes taxation, environmental policy, unionization, and voluntarily funded goods.
How it works
The problem starts with two properties economists ascribe to public goods. A good is non-excludable when it is hard or impossible to stop non-payers from enjoying it, and non-rival when one person's use does not diminish what is left for others. Clean air, national defence, a lighthouse beam, or a well-kept shared reputation all have this character. Paul Samuelson's 1954 formalisation showed why markets under-supply such goods: because each consumer benefits regardless of payment, none has a private reason to reveal how much they value the good or to pay their share. Every contributor faces the same calculation. Whatever others do, an individual is materially better off keeping their money, since they still receive the shared benefit either way. Contributing nothing is a dominant strategy, and if everyone reasons this way the good is under-provided or never appears, even though all would prefer it to exist. Mancur Olson's 1965 analysis extended the logic to organised groups and predicted, more contentiously, that large groups are worse at solving the problem than small ones: the larger the group, the smaller each member's share of the benefit, the easier it is to hide inside the crowd, and the more the outcome depends on selective incentives, meaning private rewards or penalties tied to contributing rather than to the public good itself.
What the evidence shows
The sharpest early test asked whether people actually behave as the theory predicts. In a much-cited 1981 study Gerald Marwell and Ruth Ames had subjects divide tokens between a private account and a group account, and found that most contributed between 40 and 60 percent of their resources to the collective good, far more than the zero the model implies. The one group that came close to pure free riding was first-year economics graduate students, prompting the paper's title, "Economists free ride, does anyone else?". This partial-contribution result has proved among the most robust in experimental economics. In the standard voluntary-contribution mechanism, people begin by giving roughly half their endowment, but when the game is repeated for a fixed number of rounds contributions decay steadily toward, though rarely all the way to, the free-riding prediction. Jennifer Zelmer's 2003 meta-analysis of linear public-goods experiments confirmed the pattern and quantified how design features such as the marginal return from the group account, communication, and repetition move contributions up or down. Olson's further prediction that larger groups necessarily free-ride more has fared less well in the laboratory: R. Mark Isaac, James Walker and Arlington Williams found in 1994 that groups of 40 and 100 provided a pure public good at least as efficiently as groups of 4 or 10, and Zelmer's meta-analysis does not single out group size as a reliable drag on contributions, so the large-group penalty is better read as a theoretical tendency the data qualify than as a confirmed regularity. Two explanations compete for the decay. One is that many players are conditionally cooperative: Urs Fischbacher, Simon Gaechter and Ernst Fehr showed in 2001 that a majority raise their own contribution when they expect others to give, so a few free riders drag the average down over time. The other is simple confusion about the payoff structure, which experiments that separate kindness from error find explains a meaningful part, though not all, of early giving.
Where it breaks down
Free riding is a tendency, not an iron law, and a large literature maps the conditions that blunt it. Letting people talk before they decide has a strong effect: R. Mark Isaac and James Walker showed in 1988 that even non-binding, unenforceable discussion, so-called cheap talk, sharply raised and stabilised contributions, because it lets a group form and voice a shared norm. Allowing members to punish low contributors is more powerful still. Ernst Fehr and Simon Gaechter reported in 2000 that when players could pay to impose a penalty on free riders, cooperation rose and held near the maximum contribution level instead of decaying, even though punishing is itself costly and its benefits are shared equally. That last point exposes a recursion the theory anticipates: punishment is a second-order public good, since the punisher bears the cost while the restored cooperation is enjoyed by all, so a purely self-interested actor should decline to punish too. Reputation, repeated interaction with the same partners, and smaller groups where each contribution is pivotal all push in the same direction. None of this is unconditional. Cross-cultural work has found that in some societies people direct antisocial punishment at high contributors rather than at free riders, which can wipe out the gains, so the sanctioning remedy depends on the surrounding norms rather than working everywhere by default.
Where it shows up
The same incentive structure recurs across very different settings, wherever a benefit reaches everyone but paying for it is optional and costly. Public finance is the archetype: people gain from roads, courts and defence whether or not they pay, which is the standard justification for compulsory taxation rather than a donation box. Environmental goods have the same shape at global scale, since a stable climate or a replenished fishery benefits all parties, so each has an incentive to let others bear the cost of restraint, which is why such agreements lean so heavily on monitoring and reciprocal commitments. Labour organising faces it directly: a negotiated wage rise covers every worker in a bargaining unit, so an individual may reason that the raise will arrive whether or not they pay dues or risk a strike, which is the classic argument for union security arrangements. Voluntarily funded goods such as open-source software, community-run reference works and listener-supported broadcasting survive precisely because a minority contribute while most consume, sustained by reputation, identity and small selective perks rather than by the median user's payment. Herd immunity from vaccination runs on the same logic in reverse, where a person can enjoy the protection a well-vaccinated population provides while declining the small individual cost or risk of being vaccinated.
Related but distinct
Several ideas sit close to free riding and are worth separating. Social loafing describes individuals exerting less effort on a task when their output is pooled and their personal share cannot be identified; it is a psychological cousin, but it turns on effort and accountability inside a working group rather than on the non-excludable benefits of a public good. Diffusion of responsibility is narrower still, referring to the way the felt obligation to act, to help or intervene or check something, thins out as more bystanders are present. The tragedy of the commons is the mirror image on the cost side: instead of under-contributing to a shared benefit, each user over-extracts from a shared and depletable resource, though the underlying incentive misalignment is the same. Moral hazard involves taking on more risk because someone else absorbs the downside, which shares the theme of shifted costs but not the collective-provision structure. Keeping these apart matters, because the fixes differ. Making individual effort visible addresses social loafing; assigning a specific, named person addresses diffusion of responsibility; and only the classic remedies of selective incentives, enforceable rules, or credible sanctioning reliably address free riding proper.
Examples
Using public radio for years while never donating to it.
A person skips a recommended vaccine, reasoning that if enough others are immunised the disease will not circulate anyway; if too many place the same bet, the herd immunity each was counting on never forms.
In a shared office kitchen, cleaning the coffee machine helps everyone equally, so each person leaves it for the next, and it stays dirty until a rota or a pointed note assigns the job to someone in particular.
On a peer-to-peer file-sharing network, users who download without uploading in return, known as leeching, consume others' bandwidth while giving none back, which is why many such networks enforce a minimum upload ratio.
Firms in an industry all gain from an association's lobbying for lower tariffs, so a mid-size manufacturer lets its larger rivals fund the campaign while sharing fully in any policy win it produces.
First described in Public-goods economics; Mancur Olson (1965).
Key references
- Samuelson, P. A. (1954). The pure theory of public expenditure. The Review of Economics and Statistics, 36(4), 387-389. doi.org/10.2307/1925895
- Olson, M. (1965). The logic of collective action: Public goods and the theory of groups. Harvard University Press. www.hup.harvard.edu/books/9780674537514
- Marwell, G., & Ames, R. E. (1981). Economists free ride, does anyone else? Journal of Public Economics, 15(3), 295-310. doi.org/10.1016/0047-2727(81)90013-X
- Isaac, R. M., & Walker, J. M. (1988). Communication and free-riding behavior: The voluntary contribution mechanism. Economic Inquiry, 26(4), 585-608. doi.org/10.1111/j.1465-7295.1988.tb01519.x
- Isaac, R. M., Walker, J. M., & Williams, A. W. (1994). Group size and the voluntary provision of public goods: Experimental evidence utilizing large groups. Journal of Public Economics, 54(1), 1-36. doi.org/10.1016/0047-2727(94)90068-X
- Fehr, E., & Gaechter, S. (2000). Cooperation and punishment in public goods experiments. American Economic Review, 90(4), 980-994. doi.org/10.1257/aer.90.4.980
- Fischbacher, U., Gaechter, S., & Fehr, E. (2001). Are people conditionally cooperative? Evidence from a public goods experiment. Economics Letters, 71(3), 397-404. doi.org/10.1016/S0165-1765(01)00394-9
- Zelmer, J. (2003). Linear public goods experiments: A meta-analysis. Experimental Economics, 6(3), 299-310. doi.org/10.1023/A:1026277420119
- Chaudhuri, A. (2011). Sustaining cooperation in laboratory public goods experiments: A selective survey of the literature. Experimental Economics, 14(1), 47-83. doi.org/10.1007/s10683-010-9257-1