Why Two Separate Tests Are Needed
A good is not classified simply by who provides it or whether it seems valuable to society. The standard economic classification asks two different questions: does one person’s use reduce what is available to others, and can people be prevented from accessing it? Keeping the questions separate matters. A good may be difficult to restrict access to but still be depleted by use; another may be easy to restrict access to even though one person’s use does not reduce anyone else’s.
The classification applies to a clearly specified good, in a stated setting and over a relevant period. For instance, access to a lake and rental of a canoe on that lake are distinct goods. Conditions about canoe availability do not, by themselves, establish whether lake access is rival or excludable. A sound classification keeps the good being analyzed consistent from the first test to the second.
Rivalry concerns the effect of an additional user on other users, not whether the good is scarce in an everyday sense. A good can be scarce overall yet nonrival over the range being considered: one more person may enjoy a broadcast without reducing anyone else’s reception. Conversely, a shared resource may become less available as people use it. Excludability is about practical ability to control access, not merely whether exclusion is legally permitted or whether a provider would prefer to exclude.
The Four Ideal Categories
The two tests are independent. Each has two possible answers, so together they produce four combinations. These combinations define the standard ideal categories of goods.
| Rivalry | Excludability | Category | Typical interpretation |
|---|---|---|---|
| Rival | Excludable | Private good | Use by one person reduces what remains for others, and access can be restricted. |
| Nonrival | Nonexcludable | Public good | People can share the benefit without diminishing it, and access cannot readily be restricted. |
| Rival | Nonexcludable | Common-pool resource | Use reduces what remains for others, but restricting access is not feasible. |
| Nonrival | Excludable | Club good | Access can be restricted, while one user’s use does not reduce what is available to others. |
This table follows directly from pairing the answers to the two tests. For example, “rival and nonexcludable” cannot be a public good under the standard definition: rivalry fails the public-good test, even though access is difficult to restrict. The two tests must both be applied; neither property determines the other.
These are ideal types, not claims that every real good fits one category permanently. Technology, rules, enforcement costs, and the number of users can change whether exclusion is feasible or whether additional use reduces what others receive. For this classification task, state the relevant assumptions and classify the good under those assumptions.
Worked Classifications
A reliable method is to name the good precisely, test rivalry, test excludability, and then use the table. The examples below make the assumptions explicit so that the two tests refer to the same good.
Worked Example: A Community Tornado-Warning Signal
A hypothetical town operates a siren that broadcasts a tornado warning throughout the town. Assume that every resident within the signal area can hear the warning without signing up or paying, and that one resident hearing it does not weaken the signal or reduce anyone else’s ability to hear it. Classify the warning signal.
Method: define the good. The good is the warning signal received by residents in the specified area, not the siren equipment itself and not the town’s maintenance service.
First test: rivalry. Under the stated assumption, an additional resident hearing the warning does not diminish the warning available to others. The signal is nonrival.
Second test: excludability. Residents do not need to register or pay, and the town cannot practically prevent a particular resident in the signal area from hearing the broadcast. The signal is nonexcludable.
Classification. Nonrival and nonexcludable gives a public good.
Common mistake avoided. It would be incorrect to call the warning a private good merely because the town owns the siren. Ownership of the equipment does not establish either rivalry or feasible exclusion from the warning.
Worked Example: A Subscription Research Archive
A hypothetical organization provides access to a digital research archive. Only users with valid paid credentials can open it. Assume that the archive has sufficient capacity for all users in the scenario, so one subscriber’s access does not slow or prevent another subscriber from using it. Classify access to the archive.
Method: distinguish access from ownership of a copy. The good being classified is access to the archive during the subscription period. The scenario is not asking about the sale of a physical book or the organization’s servers.
First test: rivalry. With sufficient capacity, one authorized user’s access does not reduce the access available to another authorized user. Access is nonrival under the stated conditions.
Second test: excludability. The organization can require credentials and deny access to people without them. Access is excludable.
Classification. Nonrival and excludable gives a club good.
Common mistake avoided. “Digital” does not automatically mean nonexcludable: credential controls make access restrictable here. Also, the conclusion depends on the explicit capacity assumption; without it, the scenario would not establish that users’ access never affects one another.
Worked Example: An Open-Access Groundwater Supply
In a hypothetical rural area, classify the groundwater available for extraction from one specified aquifer during a dry season. Assume that extraction by one well lowers the remaining water in that aquifer, leaving less available for other well users. Also assume that no practical monitoring or enforcement system can prevent people in the area from pumping water.
Method: keep the resource and period fixed. The good is groundwater available for extraction from this aquifer during this dry season. The question is not about access to nearby land or ownership of a particular pump.
First test: rivalry. Pumping groundwater reduces the remaining stock available to other users during the stated period. The resource is rival.
Second test: excludability. The assumption rules out a practical means of preventing people from pumping. Access to the resource is nonexcludable.
Classification. Rival and nonexcludable gives a common-pool resource.
Common mistake avoided. It would be a category error to use rivalry of the groundwater supply to infer that the well service is nonexcludable, or to use open access to infer that groundwater is nonrival. The classification follows because the problem separately establishes both properties for the same aquifer resource.
Worked Example: A Sealed Meal Sold at a Station
A station sells individually packaged meals. A buyer can be refused a meal without payment, and once a package is sold it is no longer available for another buyer. Assume the station can enforce payment and that each package is a separate unit. Classify the meal.
Method: assess the package as the good. The relevant good is one packaged meal, not the station’s dining area or the general service of selling food.
First test: rivalry. If one customer consumes the package, that same package cannot be consumed by another customer. The meal is rival.
Second test: excludability. The seller can refuse to transfer the package unless the customer pays. The meal is excludable.
Classification. Rival and excludable gives a private good.
Common mistake avoided. The fact that many packages are available does not make each package nonrival. The test concerns whether use of the specified unit reduces what remains for others.
A Short Justification for the Classification Method
Why does the two-test method produce these four categories? Begin with a specified good and ask two yes-or-no questions. The rivalry test has two outcomes: use either reduces what is available to others or it does not. Independently, the excludability test has two outcomes: access either can be restricted or it cannot practically be restricted. Pairing each outcome of the first test with each outcome of the second yields exactly four combinations.
State whether the question concerns a resource, a service, access, or a unit of a good, and specify any relevant time or setting.
Ask whether one person’s use reduces the amount or quality available to others under the stated conditions.
Ask whether people can feasibly be prevented from accessing that same good. Do not infer the answer from the rivalry result.
Rival and excludable is private; nonrival and nonexcludable is public; rival and nonexcludable is common-pool; nonrival and excludable is a club good.
The independence of the tests is the important part of the justification. Scarcity does not prove that access can be restricted; a resource can be scarce and still open to all. Likewise, the possibility of restricting access does not prove that use is rival; a provider can control access to a good that several people can use without diminishing one another’s benefit.
Why the Classification Matters
The categories help identify which features of a good may complicate ordinary market exchange. If access can be restricted, a provider may be able to limit use to selected users. If use is rival, each person’s consumption can affect what remains for others. These are different economic considerations, so a single label such as “shared” or “scarce” is not enough to describe the good.
For actuarial and economic analysis, the same discipline applies when describing a hypothetical insurance-related service or public program. Specify whether the object is a coverage benefit, access to information, a facility, or a scarce resource. Then state the assumptions about who can be excluded and whether one user’s use reduces what others can receive. Classification can clarify the setting, but it does not by itself determine the best policy, the efficient quantity, or how a service should be financed.
Common Classification Traps
- Changing the good between tests. Do not use the rivalry of a limited piece of equipment to classify access to the surrounding resource. State one good and apply both tests to it.
- Treating “public” as a statement about ownership. The category depends on rivalry and excludability, not on whether a government or private organization provides the good.
- Equating nonexcludable with free. A price might be charged in some settings, but the classification question is whether access can feasibly be restricted, not whether someone currently pays.
- Equating abundant with nonrival. A large supply can still be reduced by use. Test whether an additional user affects what remains available to others.
- Assuming an owner can always exclude. Ownership or a rule on paper does not establish that access can be monitored and restricted in practice.
- Ignoring stated conditions. Classification may depend on the relevant time period, capacity, technology, or enforcement assumptions. Use the conditions given rather than adding unsupported ones.
Check Your Understanding
For each scenario, identify the good, test rivalry and excludability, and give the category. Treat each stated assumption as part of the problem.
- A hypothetical emergency radio broadcast reaches every household in a valley without registration. Assume one household’s reception does not weaken reception elsewhere. Classify the broadcast. (Answer: Nonrival and nonexcludable; public good.)
- A museum offers access to a searchable digital catalogue only to members with active accounts. Assume the catalogue operates with enough capacity that one member’s search does not impair another’s. Classify catalogue access. (Answer: Nonrival and excludable; club good.)
- A group of unmonitored fishers can catch fish from one specified shared stock. Each fish caught reduces the stock left for other fishers, and no practical system can prevent any fisher from entering. Classify the fish stock. (Answer: Rival and nonexcludable; common-pool resource.)
- A bakery sells one loaf at a time, can refuse to hand it over until payment, and cannot sell that same loaf to a second customer after the first customer takes it. Classify one loaf. (Answer: Rival and excludable; private good.)
- A town makes a storm-warning text service available to all residents without an account. Assume messages reach all recipients and an additional recipient does not reduce the warnings others receive. Classify the service, and explain why its digital form alone is not the reason for the answer. (Answer: Nonrival and nonexcludable; public good. The stated access and use conditions establish the category, not the fact that it is digital.)