Are apples non-rival?

Suppose we had an apple replicating machine, that could replicate apples costlessly and instantly. How would we go about answering the question: "Are apples a non-rival good?".

1. We could have a lovely philosophical argument about whether the original apple and the copy are two different apples or just two different manifestations of the same apple. We would need to answer that question before we could decide whether the same apple was being consumed twice, which is one way to define "non-rival" goods.

2. Or we could simply say "yes", because the answer to the first question doesn't matter for economics; what does matter is that the marginal cost of producing the second and all subsequent consumptions of the apple(s) is zero.

Now we do in fact have apple replicating machines. We call them "apple trees". But they take labour, land, and time, to replicate apples. So the marginal cost of replicating an apple is not zero. And if we model the apple market, our model will include those replication costs. And unless we are talking about some new and rare variety of apple, we will ignore the cost of the first apple, because it's trivial.

Nearly all the food we eat is produced by replicating machines. Plants and animals are replicating machines.

If I teach you an idea, and you learn it, is the idea in your head the same idea as the idea in my head? One idea or two? It doesn't matter. What matters is the cost of replicating the idea. Sometimes that cost is non-trivial. Schools and universities are idea replication machines, and they are costly machines. They use the teachers' time and the students' time. If you start school at age 5, quit school at 20, then get a job and retire at 65, you have spent at minimum 25% of your total working years working at the job of replicating ideas. And that's not counting your teachers.

Factories that mass-produce cars are car replicating machines. First they build the prototype, then it gets replicated in the factory. The prototype is costly, and the replicas are costly too.

I can't eat the same apple twice, but I can use the same socket set twice. The first use cost me $100 (which is what I paid to buy it), but the second and all subsequent uses cost me $0 (wear and tear are negligible). I don't (normally) need to buy two identical socket sets. But two people might need to buy two identical socket sets, because the same socket set can't be in two different cars or garages (though friends and neighbours might sometimes be able to share, at little extra cost).

A live firework display is the exact opposite of the socket set. I can't watch it twice at zero marginal cost, but a second person can watch it once at zero marginal cost.

These examples show that the marginal cost of the second use may depend on whether the second user is the same person or a different person to the first user. It's the same with ideas: usually it's easier and cheaper to remember an old idea so you can use it twice, than for one person to teach another person. But at the macro level we can only remember by teaching, because individuals die. Some technologies do get forgotten, and need to be rediscovered.

Ideas are economically important, but are they economically special, on the rivalry/non-rivalry dimension? Even boring material goods, like socket sets, can get pretty weird, if you start thinking about them.

[Update: jonathan makes a good point about excludability of replication of ideas in comments. If I sell you an idea by teaching it to you, you also (usually) have the ability to teach it to others. That's not true if I sell you a car (unless reverse-engineering is easy). It's true if I sell you a live cow for milk, but not if I sell you a dead cow for meat.]

Thanks to commenter jonathan (who is not responsible for opinions here).

[Paul Romer might like my old post "Macroeconomics when all goods are non-rival", (even though it's short run macro and not growth theory like he does). And I was doing (short run) macro with monopolistic competition (pdf) (a couple of months) before it was fashionable. So it's especially silly (not to mention intolerant of disagreement) for him to make this accusation of me: "After twenty years of playing whac-a-mole with the die-hard price-taking neo-Marshallians, I can tell you exactly where the argument will go next. When the choices are stated as clearly as Dietz does, the neo-Marshallians will claim that there is no such thing as a nonrival good. Nick Rowe has already started laying down a smoke screen of obfuscation on this point." Nope. He got that one totally wrong, on both counts.

And that sort of response doesn't create a productive environment for new ideas. People need space to muse over old and new ideas, without the authorities making nasty accusations about their motives.]

Comments

[Joshgans]

The apple machine proves Romer's point. Suppose to produce apples you needed that machine then is competitive provision possible? If you pay all factors of production their marginal products, there is nothing left to pay for the machine. The same is true for ideas. Thus, the issue with non-rivalry is that you can't get a competitive equilibrium that pays for all production.

And this is the point. This happens all over the place. That is why perfect competition doesn't actually arise in reality. So why would we care about questions of whether it can arise?

[Nick Rowe]

Joshgans: you are arguing about my older post: http://worthwhile.typepad.com/worthwhile_canadian_initi/2015/06/mathiness-and-growth-theory.html where I gave a counterexample to what Romer said.

As Dietz Vollrath said: "There are really two questions we are dealing with here. First, do inputs to production earn their marginal product? Second, do the owners of non-rival ideas have market power or not? We can answer the first without having to answer the second."

If there are taxes, for example, then you cannot pay all factors their marginal products. But taxes don't necessarily mean that individuals don't face perfectly elastic supply and demand curves.

This post is not about whether the economy is perfectly competitive. It isn't. Nor about whether there are IRS technologies. There are.

[Jeff]

@Joshgans, Yes, this! (Totally agree.) And as to "why would we care about questions of whether it can arise?", the answer is because a lot of arguments start with "suppose we have a competitive equilibrium", just as Nick's post here starts with "suppose we had an apple replicating machine", and David's paper (probably) says "suppose we have a free education system already in place" (in so many words). In short, because people keep trying to get away with "as if" arguments and forgetting there was even a "suppose" at the beginning (untethering). Romer seems to be saying "it's intellectually unproductive for [him] (or anyone) to have to keep reminding the authors of all the 'as if' bases of their arguments" (whac-a-mole).

@Nick, It's a little dishonest to claim you are not arguing against non-rivalry on the basis that "here I am saying 'even apples may be non-rival'". I read you as attempting a reductio ad absurdum (on the rival/non-rival distinction) in the post. Am I wrong?

[Alex]

Nick, Please change the topic. Stop trolling Paul Romer. He doesn't deserve it. Thanks.

[Nick Rowe]

Alex: sorry. Too late! The whole rival/non-rival thing is now stuck in my head like a catchy pop song! And I've got to get it out, regardless of Paul Romer. (And I've just re-read my old post on Macroeconomics when all goods are non-rival, and still like it (apart from my use of the metaphysical definition of non-rival), and am wondering if it deserves a follow-up.)

Jeff: I am arguing against *one definition* of the rival/nonrival distinction. And I am arguing that the rival/non rival distinction has (at least) two dimensions: across individuals (live fireworks); across multiple uses by the same individual (socket sets). I am arguing there's nothing special about ideas on the rival/non-rival dimension.

But the idea that I am a die-hard price taking neo Marshallian claiming there is no such thing as a non-rival good is just silly (though rivalry is more a matter of degree, with pure non-rival being one extreme).

[jonathan]

Another issue is whether it is possible to "sell" the good without also selling the replication technology.

If I buy a car, I don't thereby gain the ability to manufacture that car. I could reverse engineer it, which would be a lot easier than inventing a car from scratch, but a lot more difficult than reading off the blueprint. But if I learn the Pythagorean theorem, I gain the ability to teach it to others. Thus there's a sense in which the design and the output from that design are separable in the case of the car, but not in the case of an idea.

If ideas were like cars in this sense, the inventor of a theorem would be the only one who could teach it. Or at least, if someone wanted to be able to teach it, they would have to expend a lot of effort beyond simply learning the idea themself.

[Nick Rowe]

jonathan: "Another issue is whether it is possible to "sell" the good without also selling the replication technology."

Good point. Certainly affects excludability (absent legal restrictions, of course.) Depends if the cow you buy is alive or dead, or if the seeds have been processed so they won't germinate.

[JR Hulla]

This whole argument may become economically significant as 3-Dprinting costs plummet at the same time as material properties increase, not to mention 3d scanners making it easier to copy complex shapes. What happens to the entrepreneur who designs an indispensable $3.95 gadget? If large number of people have access to scanners and 3d printers (some libraries have them) there is no practical way that one could afford to enforce any form of intellectual property protection. How about people just e-mailing the file for such a device to their friends? And just wait until desk-top gene hacking really gets going…...http://www.wired.com/2007/12/start-hacking-l/ Who was that old guy who talked about owning the means of production?

[Nick Rowe]

JR: My sense is that it has already been happening, for the last 100 years, with manufactured goods. (Apparently, "manufactured" used to mean "hand-made"). And you could argue it's been happening for a few thousand years, with agriculture.

[anon]

I think the interesting difference between socket sets and usual public goods (including ideas) is this: the buyer of a socket set generally internalizes the benefit from all uses, with no need for market transactions. Some firework shows are also like this: they're paid for by large(-ish) organizations, which pursue the interests of prospective viewers. This old post by Frances discusses another interesting case. When benefits are non-excludable and Coasean transactions are too costly, it's hard to make sure that things like ideas are efficiently rewarded.

The price-taking/marginal-cost issue looks like a bit of a red herring, at least to me. We know how fixed costs can be defrayed efficiently via Ramsey pricing, which is approximated closely enough in monopolistically-competitive industries. But this has little to do with _non-excludability_, of course.

[Luis Enrique]

Yes Romer's conduct is doing some damage to his reputation. Although - ICYM my comment on other thread - somebody on twitter pointed out to Romer that he'd misread you, and he seemed to accept that.

[Nick Rowe]

anon: " We know how fixed costs can be defrayed efficiently via Ramsey pricing, which is approximated closely enough in monopolistically-competitive industries."

The second part of that sentence doesn't seem right to me. Suppose we have an economy of monopolistically competitive firms. Start in equilibrium. Price is above MC, and so Marginal Benefit is above MC. Hold the number of firms fixed, hold prices fixed, then loosen monetary policy so the AD curve shifts right. Unless the economy-wide MC curve is vertical, all firms expand output and welfare increases. (That's why welfare increases temporarily in a boom, according to New Keynesian models.)

Thanks Luis. I didn't see that bit on Twitter. But it pisses me off when I can't post my random musings about my own confusions and my attempts to get my own head straight on things without some guy accusing me of deliberate obfuscation (to try to preserve the price-taking model!!, and when I was doing macro with monop comp before almost anyone else!!).

[anon]

Nick, that's not a ratexp equilibrium though, is it? By loosening monetary policy and keeping nominal prices fixed we're essentially pushing (real) P closer to MC, which means that quasi-rents, fixed costs etc. are not being defrayed to the same extent. So this cannot be an expected outcome.

Of course in the real world, an increase in AD might sometimes obviate coordination problems and lead to a stable improvement, as in your "macro model with all goods being non-rival". But this is not a given, and in any case such pervasive coordination failures are very hard to solve either with markets/transactional solutions, or through government. They're basically not on the radar at all.

[Nick Rowe]

anon: no, it's not (usually) a ratex equilibrium to have a boom. But the representative agent is better off in a boom, which shows that the ratex equilibrium is not efficient. And firms' profits may increase or decrease, depending on the elasticity of the economy-wide MC curve. For an extreme example, take my little macro model I link to in the post, where all goods are non-rival, so extra copies always have an MC of zero. In a boom, all firms increase the number of copies sold, and the relative price of copies stays fixed at one, so each firm earns higher profits.

[Jeff]

@Nick 05:11 PM, Arguing against a definition? Arguing for multiple definitions? How is that NOT obfuscating (and/or Strawmanning) as viewed from the other side?

The only question should be, does Romer's argument go through with the definitions HE is using or not. (Answering that clearly and unequivocally before moving on.)

Don't get me wrong, discussing alternatives is also a good thing, and I've learned a helluva lot because of this controversy. But I've also seen it get inappropriately political, as Romer warned. (Criticizing Romer's "reputation" rather than his argument, and David A. seemingly sitting on my latest reply to his blog, e.g..Wagon-circling IOW).

[William]

I might tend to disagree about the cost of teaching an idea.

A university may be expensive to attend and run, but that is because it is a system for teaching many thousands of individual ideas to a (relatively) small number of people at a time. Each idea in a lecture may take only a few minutes to explain, in itself the cost is nearly negligible. But that very small teaching cost, times all of the hundreds of thousands of things you learn over 20 years of school is where the cost comes from.

Indeed, if we use a different method of spreading the idea (say, a blog post), then while there is some cost to typing up the post, from that point on like any other intellectual property it can be replicated and learned by an infinite number of people at no marginal cost (other than the electricity to load the page). In fact, someone who doesn't even understand the idea can replicate it once it is written down.

[Jeff]

@William, Good point. Nick was looking at it as an idea provider (professor), whereas Paul R. is working on the idea consumer side (theory of production). And as anyone with work experience in industry knows, very very few of the ideas we "crammed in" during our education get incorporated into our work "production", and the ones that do are SO leveraged by repeated use that even any small cost they DO have is amortized practically to zero. Meaning: modeling ideas as zero-MC/zero-MP (non-rival) on the production side is entirely justified as a practical matter. (I interpret Paul R.'s characterization of Nick's counterarguments to mean he is frustrated by the non-self-awareness of the makers of these counterarguments as being of type "so much angels on a pinhead".)

[Frances Woolley]

Nick - this reminds me of that old "characteristics" literature which attempted to break down goods into "characteristics" that provided utility directly. So e.g. food would have a variety of characteristics like calorie content, protein content, taste, and those characteristics would directly yield utility.

So "ideas" (or calories or protein or whatever) are characteristics embedded in goods and services. But technological change has radically altered the way the goods and services which hold these characteristics are produced and distributed. E.g. some song about a bicycle can be put on a piece of vinyl or a cassette or a CD which has consumption characteristics very similar to that of your socket set (and, likewise, can be shared between friends), or it can be performed live, with consumption characteristics very similar to the fireworks display, or it can be posted on youtube, where it can be listed to over and over again by many people.

I guess this was the point you've been trying to make for a while, but I've only just worked it out?

(p.s. ignore what Alex says).

[baconbacon]

The socket/fireworks examples fail to get to the core differences. A person can teach an idea without losing that idea themselves. If you loan or give a socket set you now have 1 less socket set in your possession, loaning or gifting comes with the opportunity cost of keeping it for yourself. Teaching an idea carries no such cost (and for some even enhances their own understanding), it is this distinction that defines goods on the rival/nonrival spectrum.

[Frances Woolley]

baconbacon - Walden.

[baconbacon]

Frances Woolley- I dont follow.

[Frances Woolley]

baconbacon - Thoreau wrote a book called Walden Pond, which extolled the virtues of sharing socket sets (or perhaps it was an axe, I'm not quite sure) with one's neighbours. The idea, IIRC, is that the neighbour takes good care of the socket set, and returns it in better condition than she received it in. So sharing doesn't have costs - in fact it generates benefits - for the lender.

[baconbacon]

I don't see how that relates to the Rival/non rial discussion

[Jeff]

As to the very last line in the post, regarding motives: Romer made his position clear, at least in the Ferguson case, as to "intent": "Whether words are misleading is a question of fact, not intent, and should be treated as such." That's not very politic, but politics (in academic economics writing) is precisely one of the things against which he is railing. It would be hypocritical of him to be politic.

[Nick Rowe]

William: an idea that's just floating around in the aether is of no use to anyone. It has to be in someone's head to be of any use (or maybe in a computer??). And it's costly to get it into people's heads. Reading a book or a blog post takes time. The MC is not just the paper or electricity. If you consider all the resources that go into teaching and learning, not just schools and universities, but parents and kids, and on the job training, my (wild) guess is that teaching and learning takes up somewhere around 50% of GDP (if we counted all the things like parents training their kids that normally don't get counted in GDP).

[Nick Rowe]

bacon: the second eating of an apple has a marginal cost that is independent of whether the second eater is the same person as the first eater. The second use of a socket set has zero MC, if the second user is the same as the first user. The second watching of a firework display has zero MC, if the second watcher is different from the first watcher.

[anon]

Nick, but my point is that your model shows a pervasive coordination problem (as is generally true of the elastic-MC case). It can't be a simple mispricing issue (as in the case of a non-price-taking, monopolist firm), because any _single_ firm cannot lower its price and still pay for its inputs. There are no excess returns in monopolistic competition, in contrast to monopoly; any excess rents should be competed away via new entry. Overall, I'm not sure why folks are talking about this huge issue that paying every factor its marginal product and pricing everything at marginal cost leaves no room for 'non-marginal' things like ideas, because of Euler's theorem. In the real world, we're obviously going to see some deviations from "perfectly competitive" MC pricing _and_ MRP compensation, but overall we know what these are supposed to look like, and the effects are not always going to be significant.

[Longtooth]

The argument made: "If I teach you an idea, and you learn it, is the idea in your head the same idea as the idea in my head? One idea or two? It doesn't matter. What matters is the cost of replicating the idea."

If the idea you tried to pass to another is not replicated identically in the other then you didn't replicate the idea at all. Hence the 'cost' you describe was spent for naught, a waste of your time, hence a waste of resource. And then there was no replicative costs involved. Hmmm.

You assume, for the purposes of economic discourse that that an idea has a replication cost if and when it is passed to somebody else whether or not what was received is in fact the idea you actually tried to pass.

I think, but am not sure what you are trying to say is simply that human communications have a cost of time and effort... incurred by each party in the communication. However, in economics, those costs only have relevance if the communication ever results by even the remotest relation to what was communicated in a product(something that can be traded at a non-zero value or adds value).

This cost must then be concluded to be the costs of human experiences over time, since idea's which result, ever, in any value add are the result of both the accumulation of observations and experiences, of which communication is but one form of experience. It appears from your 6/19 @ 8:25 post that you're confirming my take on your premise: to wit: all things have marginal costs, so that the only economic issues are their magnitudes and how they are allocated.

From that then in the non-rival/rival debate there can be no difference between them .. both have marginal costs, both must therefore be allocated, so there is then no economic difference on that basis. Since there is no economic difference there then non-rival = rival, which is to say there's no such thing as a non-rival good.. unless you also want to say there's no such thing as a rival good.

I've noted that in most posts on the subject, there's this assumption that effectively describes some costs as "negligible" in different forms of different types of replication.

Thus one problem in the debate is that nobody's willing to objectively define the line between "neglible" and "non-negligable". In your socket set argument, as one poster pointed out quite nicely, there is an opportunity cost to lending the socket set to somebody else.. and there's also the wear/tear cost... so somewhere in this debate the lack of objectivity is lending to further obfuscation of the real question.

I fail to see how the issue can be answered objectively when both sides of the debate refuse to engage in objective argument. Thus far, reading from the very beginning of the original back and forth several days ago this is purely a "he said/she said" debate. Interesting to me that economists who like to describe themselves as pursuing a form of science or at least not a form of theological belief systems, are so inept at objective analysis methods. Make me wonder wtf they're teaching in schools of economics.

[Nick Rowe]

anon: I'm not following you there. And since you clearly know your stuff, I want to try to resolve this. Let me restate my point, then ask you to restate your point.

Take my model where all goods are non-rival. Start in Long Run equilibrium (assume it exists, is unique, and is stable). That is a "Long Run" equilibrium in two quite different senses of the word:

1. The "Micro" sense: zero profits and no incentive for firms to enter or exit.

2. The "Macro" sense. We are on the vertical LRAS curve, where firms have had long enough to adjust their money prices, and none wants to change its relative price.

Now: hold the number of firms temporarily fixed, and hold nominal prices temporarily fixed, (so we are in the short run in both the micro and macro senses), and increase the money supply. There is a (first-order) increase in both profits and welfare, as output (the number of copies) increases. I call that a "boom". (The opposite is a recession).

Firms now have an incentive to increase prices (because I have assumed unique and stable LR equilibrium), and new firms have an incentive to enter. Eventually the price level rises, and we go back to the LRAS curve, with the same number of firms, and the same output and welfare, as originally. (I tend to assume that prices adjust more quickly than new firms enter, so we reach macro LR more quickly than we reach micro LR, but I'm not sure that assumption matters for our disagreement here.)

Over to you.

Longtooth: "From that then in the non-rival/rival debate there can be no difference between them .. both have marginal costs, both must therefore be allocated, so there is then no economic difference on that basis."

No. There is a difference in degree, and it's an economically important difference. "Pure/strict non-rival" is the extreme limit of a continuum, where MC for the second and all subsequent uses is zero. But "ideas" are not (generally) at that limit, and sometimes not even close. (There may or may not be any real world goods at that extreme limit, though that doesn't mean it isn't sometimes a useful simplification to build a model which assumes there are goods at that limit.)

[baconbacon]

@ Nick-
The first person to have an idea is analogous to the person putting on the fireworks disply, not the first watcher of the display. To produce a display he must destroy some number of his store of fireworks. A non rival good means the original holder of the good still owns/possesses that good after he has bestowed it on another person. A fireworks display is non rival for individuals in group X where X is thenumber of people that can see the display, but is rival between all possible groupings of X. Shifting the launcher 10 ft to the right means (with perfectly distributed watchers, on level ground with identical eyesight, yada yada) group X shifts in composition slightly to X'. The MC for X+1 is very high (an additional fireworks show).

Even granting a show that is visible to the entire human population the MC is that of an entire show for the group X. For a better example- the marginal cost of the second person getting into my car and heading to the same desitination as I is very low, up till the capacity of my car, at which point passinger X+1 needs a entire vehicle extra. Ignoring the costs of carrying mroe weight my sedan could be said to be non rival over the interval X where 1<=X<=5 (but only in a limited way since the desitinations of the 5 people are heavily limited).

As far as non rival goods go- let me propose language as very close. A child will learn language with basically no effort from their parents as long as their parents (or basically anyone) are using language near the child.

[baconbacon]

It seems I could have been way clearer.

A fireworks display is rival through time. It can be viewed at time t or at time t+1 or at time t+2 etc but it cannot be viewed at time t and then at time t+1. The marginal cost of watching at t+1 is the entire cost of the fireworks display. An idea can be used at time t, and then at t+1 and then at t+2 etc. Instead of thinking of the cost of teaching think of an economy of 1 person. He can use an idea over and over again without consuming that idea. "Its raining, I will stand under this shelter to keep dry" can be used every time that it rains, and the marginal cost of producing the idea the second time is basically zero.

[Nick Rowe]

bacon: that's a good way of putting it. Yes. A firework display is rival across time, but non-rival across people. A socket set is rival across people, but non-rival across time. An idea is non-rival across both people and time. (Assuming no costs of teaching and remembering).

[anon]

Nick, I don't dispute what you say about your monopolisticsly competitive model - I think that's correct. I only have a problem with people who say that prices being set at the margin is so important to a definition of perfect competition that a "competitive" model cannot possibly say anything interesting about ideas (Because of Euler's theorem). I think that the math is leading some people astray and making them miss the basic commonalities that can exist between perfect competition and more flexible models that do allow for some degree of market power.

[djb]

Joshgans,

agreed, much of theoretical economics focuses on "pure" mostly unrealistic scenarios

that was genius of Keynes, in my opinion, his theoretical concepts work in the real world

[Odie]

Absolutely interesting series of posts. Maybe I can add some thoughts of mine:

How about comparing ideas with shapes? An apple is a certain manifestation of matter and energy in a specific shape. That's why we call it an apple and consider it a food. An idea also requires matter and energy, specifically an arrangement of neurons in our brain. The matter in the brain and the apple are rival goods, they cannot be in the same place at once. The "shape" they are in is non-rival. One item having a certain shape does not exclude another item from having the same.

We call it "producing" when we form matter into a new shape using labor. We call it "learning" when form (brain)matter into a new shape again using labor. The final product of learning we call knowledge. When we "consume" the apple we transform that certain shape of the matter into other shapes; mostly individual molecules. When we die we transform our knowledge into the same. In that sense, it also gets consumed. Knowledge has a rival and non-rival component as does an apple. Producing both can be repeated essentially to infinity as the amount of energy and matter stays the same as long as we can supply the labor to do so.

[Jeff]

Odie, thanks for sharing those thoughts, as they connect with ideas I know from a different domain. (Although now we're getting deep into philosophical potential fantasy-land (as far as economics goes).) Robert Pirsig, author philosopher, called what you're talking about "patterns", and he usefully separated them into types (levels) as you are doing here. An apple, as food, is a pattern of atoms and molecules (physical level), as well as a biological-level pattern that "packages" those into a unit (covered by the skin, etc.). Picking and eating (biting, chewing) breaks the biological pattern apart (kills the organism), while digesting alters the physical pattern (changes and disperses the chemicals). It's the patterns (of "apple", of "sugar molecule") that are non-rival, as you say.

"Producing" in economics is seemingly defined as altering physical-level patterns (chemical processes, building stuff), as well as, on the agriculture side, destroying biological patterns (killing, cutting up, and packaging living things).

"Learning" deals in patterns in the separate "intellectual level". Again, the patterns here (ideas) are non-rival.

Pirsig also identified a "social level" that doesn't get talked about in economics much. But this is where the "politics" to which Romer refers is a pattern. Government and language are other patterns on this level I think. All four of these levels fit together in a hierarchy with competing values and priorities, intellectual>>social>>biological>>physical.

Pirsig's "base of operations" is an insight he had about what he calls "Quality" which encompasses "values" of all types, which is where economics may come in.

He also identified another overriding "Dynamic" level, which is where I think the "creative destruction" idea fits.

Anyway, hopefully more food for thought.

[Eric Lonergan]

Some software has zero cost of replication but rivalrous - I can't use yours. National defence has high cost of replication, but non-rival. When cost of replication is near-zero, rivalrousness doesn't matter ... Unless the machine breaks down ...