I get why, say, the military is a public good, but I don't get why public roads are public goods, since they can't be non-rivalrous, given that at any time only a subset of the population can actually be on them.

  • 1
    $\begingroup$ This question and answer may be helpful: economics.stackexchange.com/questions/12520/… $\endgroup$ Commented Apr 27, 2020 at 9:46
  • $\begingroup$ Who claims that roads are public and not just common goods? $\endgroup$ Commented Apr 29, 2020 at 8:25
  • $\begingroup$ @I'mwithMonica No one claim that. The claim is "all public roads are public goods". It's not equivalent to "all roads are public and all roads are public goods". $\endgroup$
    – Atonal
    Commented Apr 30, 2020 at 2:31
  • $\begingroup$ @Atonal, sorry if my question was ambiguous. What I meant was: Where does your claim, that "all public roads are public goods" come from? I have never heard that claim before. "Public" roads are common goods. ("Public" in this context is not an economical term!) - Thus, the premise of your question seems to be flawed: They are not, period. $\endgroup$ Commented Apr 30, 2020 at 7:30
  • $\begingroup$ @I'mwithMonica Ah, that makes sense! I first understood that to be the case here (mason.gmu.edu/~atabarro/PrivateProvision.pdf). Mind you that I don't have an Econ background, and I did not document my cursory review of internet resources! But now following, your questions, this "Samuelson and Nordhaus 1985: 48-49" seems to be an example (Found here: marketurbanism.com/2011/02/22/urbanism-legend-public-good) $\endgroup$
    – Atonal
    Commented Apr 30, 2020 at 10:34

2 Answers 2


The property of rivalry is a continuous (rather than binary) variable.*

A good is rivalrous if my consumption of it reduces the amount that can be consumed by others.

So, a particular Big Mac is fully rivalrous, because each bite I take from it reduces (by that exact same amount I've bitten) the amount left for you.

The degree to which roads are rivalrous is lower: My driving on a deserted road at night does not reduce at all the amount that can be driven on the same road by others.

But your doubt is correct—it is not generally $0$: My driving at peak hour at a city center does reduce the amount that can be driven by others.

Ideas are a good (and possibly sole) example of a good that's perfectly non-rivalrous. (My "consumption" of the Pythagorean Theorem leaves no less for anyone else.) Side point: But with the right laws and accompanying enforcement—e.g. patents, copyright—even ideas can be made excludable.

Defense is often given as a classic example of non-rivalry. But I would argue that although defense exhibits a very high degree of non-rivalry, it is less than perfectly non-rivalrous. (The resources used to defend John in New York from a nuclear attack does reduce the US Department of Defense's ability to defend Jane in Hawaii. The resources used to defend a nation with over 1B people are greater than the resources used to defend a nation with less than 1M people.)

*Likewise, the properties of excludability and being a public good are continuous (rather than binary) variables. (Many introductory textbooks tend not to emphasize this important point sufficiently.)

  • $\begingroup$ Thanks! I have some questions though. 1) When economists say something is a public good, are there saying there's a subjective threshold for the property of rivalry and excludability where the good is under them, or are they just not being rigorous enough? $\endgroup$
    – Atonal
    Commented Apr 27, 2020 at 1:58
  • 3
    $\begingroup$ (1) Again, it's a continuous variable. So this is a bit like asking when we can say that someone is tall. There will be clear-cut cases where everyone agrees that someone is (or isn't) tall, but there'll be plenty of cases where reasonable people can disagree. (2)(a) Resources used to defend any potential attacks on NY could instead have been used for Hawaii and vice versa.(b) If I drive on a deserted road, I do not at all reduce anyone else's usage of the road (since there is literally no one else using the road). $\endgroup$
    – user18
    Commented Apr 27, 2020 at 2:07
  • 4
    $\begingroup$ The purpose of discussing public goods is to argue in favor of their provision by the public, i.e. the government (hence the name). To the degree that a good is non-rivalrous and non-excludable, it is a good candidate for public/government provision. But these may not always be the sole or even decisive factors. Similarly, to the degree that someone is tall, he is a good candidate for playing basketball, but again, this may not be the sole or even decisive factor. $\endgroup$
    – user18
    Commented Apr 27, 2020 at 2:13
  • 8
    $\begingroup$ My driving on a deserted road at night does not reduce at all the amount that can be driven on the same road by others — even there it's not fully zero, because you're contributing to wearing out the road? $\endgroup$
    – gerrit
    Commented Apr 27, 2020 at 8:35
  • 1
    $\begingroup$ @gerrit Sure, but that's being a bit pedantic... The whole point of Kenny's answer is that it's a continuum. $\endgroup$
    – Josh Eller
    Commented Apr 27, 2020 at 12:16

It depends actually a non-congested road can be considered non-rivalrous as when you drive there you don’t really reduce the enjoyment or marginal utility of other people driving them.

However, once the road is congested it becomes rivalrous.

Also note that even though many textbooks include public roads as public goods this is actually not technically correct. Technically since roads can be protected by toll booths they are excludable so it’s not correct to categorize them as public goods, they are actually a separate category of quasi-public goods - that is goods that have many characteristics of public goods but also some characteristics of private goods.

However, it’s very hard to find proper example of pure public goods (although army is a good example, non-polluted air another), so often undergraduate textbooks that are less nuanced than graduate textbooks include quasi-public goods that are very close to public good in same category.

  • $\begingroup$ Makes sense. However on non-polluted air being a pure public good, couldn't you say the more people there are, the less oxygen would be available for everyone else?! Specifically, given average O2 consumption of a human, there exist some number of humans where the global rate of oxygen consumption would be more than the rate of its production? $\endgroup$
    – Atonal
    Commented Apr 27, 2020 at 1:51
  • 2
    $\begingroup$ @Atonal yes it’s always all about context there is no worry that we ever run out of O2 at any realistic human population growth. Of course O2 at the international space station is neither non-rivalrous nor non-excludable. Same under water. However, on earth at and above sea level it’s one of the best example of pure public goods. $\endgroup$
    – 1muflon1
    Commented Apr 27, 2020 at 1:56
  • 1
    $\begingroup$ @1muflon1 I don't think your examples of pure public goods really hold. Because everything is quasi-public. You can totally have an army that only serves/protects one person/one group of people. That would be similar to the toll booth. Same goes with the non-polluted air. Air pollution makes the non-polluted air rivalrous (if I pollute air, it is not non-polluted for you any more) and then having e.g. fine-dust filters in your buildings gives you non-public clean air. $\endgroup$
    – Dakkaron
    Commented Apr 28, 2020 at 7:08
  • $\begingroup$ If you think the last statement is exaggerated, have a look at how things are handled in heavily polluted cities around the world. $\endgroup$
    – Dakkaron
    Commented Apr 28, 2020 at 7:10
  • $\begingroup$ Useful information is a pure public good. $\endgroup$
    – beroal
    Commented May 7, 2020 at 18:39

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.