The Library · EconomicsPlate № 455 · Folio VI
ILL. № 455
ECON
Plate — Tragedy of the Commons

Tragedy of the Commons

Hardin 1968: shared resources collapse under individually rational use. Ostrom 1990: communities govern them stably, no state or privatisation needed.
Facets
  • Hardin's pasture and the inevitable over-usenot yet tested
  • Ostrom's design principles for self-governed commonsnot yet tested
  • Where Hardin still holds: atmosphere and open oceannot yet tested
The brief

Garrett Hardin's essay The Tragedy of the Commons (Science, 1968) argued — using an English-village pasture as parable — that a common-pool resource (rival in consumption, non-excludable) is doomed to over-use: each herder gets the full benefit of one more cow while the cost of overgrazing is shared and the pasture collapses. Hardin emphasised coercive regulation or changes in property arrangements and gave little attention to community self-governance. The empirical correction came from Elinor Ostrom, whose Governing the Commons (1990) documented communities sustainably governing common-pool resources without either — winning her the 2009 Nobel in Economics, the first awarded to a woman.

Hardin's logic is airtight given his assumptions, which is exactly why the assumptions deserve the scrutiny. A herder who adds one more cow captures the whole gain from that cow and bears only a fraction of the cost of the thinner grass, since the thinning is shared out among everyone. Every herder faces the same arithmetic, so every herder adds, and the pasture fails — not through greed but through a structure in which the private calculation and the collective one point in opposite directions. Ostrom's contribution was to identify which assumption carries the weight. Hardin's herders are strangers: they cannot see what the others take, cannot talk, cannot punish. Real commons are usually nothing of the kind. When the users are a known and lasting group, taking becomes visible — and once it is visible it can be answered, with a word first, a fine later, exclusion in the end. What she found across hundreds of cases was not one institution but a recurring shape: a boundary saying who is in, rules cut to the local resource rather than imported wholesale, the users themselves writing those rules, monitoring done by people with a stake in the answer, penalties that start small and escalate, and an outside authority willing to recognise the arrangement instead of overriding it. The Maine lobster fishery shows the shape working. State law sets size limits and protects breeding females; the harbour gangs, with no legal standing at all, decide who may set traps where. Neither half would hold alone, and together they have kept the fishery productive for more than a century. The correction has teeth because the original framing was acted upon. Read as proof that a commons must be either privatised or nationalised, it licensed enclosures and collectivisations that dismantled functioning local arrangements and left the resource worse governed than before. Where Hardin still bites is where his assumptions come true. The atmosphere has no boundary, no visible taking, and no way to exclude a defector — which is why climate agreements keep reaching for pledges and review cycles, trying to manufacture the visibility a village pasture has for nothing.

Why nowFisheries are the most-studied modern commons: individual transferable quotas (ITQs, the privatisation route) have had mixed success — New Zealand's hoki recovered, Iceland's cod is contested — while co-management arrangements (Ostrom-style) have outperformed both pure-state and pure-private regimes in many small-scale fisheries. The global atmosphere is a Hardin-shaped commons, and the question is whether nested governance (national targets within international agreements within voluntary city pledges) can substitute for the central authority Hardin insisted was necessary. Antibiotic stewardship is a slow-motion commons tragedy; groundwater in California and the High Plains is a drawdown problem where Ostrom-style governance often fails; AI training data is becoming a commons in the technical sense.