Good article Brian. I get the feeling that we use so many models in intro micro that we confuse people as to the basic message. My only gripe with this piece is that you do use the word competitive without being too precise what you mean. You seem to mean something like, we reach the same intersection of the two curves, but I still want more explicit standards as to what would make a market "competitive". What kind of market would you call non-competitive? (Presumably something with high barriers to entry for firms). Anyway, good stuff. Big fan. When I teach micro I'll certainly be using some of these with students.
Thanks for the comment. I should have been more explicit.
When we draw supply and demand we are assuming people are price-takers. That’s same for a Walrasian equilibrium.
The complication is that we can always assume people are price-takers. But should they? So a competitive equilibrium is when everyone “should” be a price taker or would act as if a price taker, even if given the choice. So Bertrand is competitive since the equilibrium is the same as one where we forced everyone to be a price-taker.
Wanting to be a price-taker turns out to mean that each individual faces a perfectly elastic curve, at least locally. So unlike the standard monopolist with downward sloping demand, the perfect competitor can’t withhold goods to raise the price.
Does that make sense? We can always complicate things so that the competition is along a different dimension, such as competition for an exclusive contract compared to competition for the final marginal units sold.
Even though I agree with part regarding asymmetric information, I'm wondering if asymmetric information could come to trouble, especially in healthcare sector. Or should it be called "asymmetric chance" since patients have much less chance to prepare for health-related events, even more so in an accident. I believe you've heard this question a lot from your students but it'll reall help if you could enlighten me a bit. Thanks in advance!
It's an important question. Absolutely. We need to worry about it. The claim is not that any asymmetric information is fine. Some is. Some isn't. Some is compatible with the perfect competition model. Some is not. We need to dig in and be more specific in the case.
I agree with one of your conclusions, but I disagree with your proof. So I want to quibble a little.
You say that profits can exist under perfect competition because of the existence of producer surplus. But producer surplus is really profit over marginal/variable costs, ignoring fixed costs.
Producer surplus is the difference between the market price and the price the firm would have accepted. In the short-run, the firm will accept any price that is greater than its variable costs, even if it fails to cover its fixed costs. So the existence of producer surplus may merely indicate that firms are covering their variable costs, not that they are making profits (i.e. revenue exceeds total costs).
However, we can still prove your proposition in another way, by using the firm side rather than the market side. I.e., draw the perfectly competitive firm's graph, with horizontal P=MR, upward-sloping MC curve, and U-shaped ATC curve. The marginal firm will have P=ATC and make zero profits. But other firms producing an identical product at lower costs will earn the same P=MR but have a lower ATC, creating profits (P>ATC). Thus, the marginal firm makes zero profits, but all infra-marginal firms will make positive profits.
In my Econ 101 course, I use the example of growing avocadoes in Alaska. If the demand curve for avocadoes shifts right on the market side, then the P of avocadoes increases. All avocado farmers temporarily earn profits. In the long-run, supply increases, and the price falls back down until their profits disappear.* But here is where I make the next distinction: suppose that the way we increased the supply of avocadoes was by building greenhouses in Alaska. This clearly costs more than growing avocadoes using sunlight in California. So the Californian farmers will earn Ricardian rents. The marginal Alaskan farmer will earn zero profit because Alaskan avocado farmers will build greenhouses up until the point where the last greenhouse earns zero profit. But at that point, Californian avocado farmers are selling identical avocadoes with a lower ATC because of the free sun and warmth of California. So the marginal Alaskan avocado farmer earns zero profit, while the Californian avocado farmers all earn Ricardian rents.
I tell my students that the reason why profit disappears is that profit creates an incentive for competitive entry. If all firms are identical, then they all compete the price down to the cost of production. But if firms are not identical, then instead, firms enter until the last firm earns zero profit. Prior firms with lower costs will be earning profits. But the final firm that enters will earn zero profits, and no one else enters after that.
* (Aside: if it a constant-cost industry, then the long-run supply is perfectly elastic and price falls back down to its original level. If increasing-cost, then the long-run supply is upward-sloping - but more elastic than short-run supply - and the price falls back down only partway towards the original price. But either way, if all avocado farmers are identical, then in the end, all profits return to zero. Identical firms plus increasing-cost implies that costs for all firms increased identically. Perhaps increasing the supply of avocadoes caused an increase in the demand for fertilizer, so fertilizer prices increased equally for all farmers.)
I could have been more explicit about assuming fixed costs being zero. But I’m not sure why you want to import all the baggage of Viner’s curves and fixed vs. variable costs. Sure, that’s a story of how markets can get more competitive over time. Sure. But it’s not the definition of perfect competition. Even Mankiw makes this distinction.
I don't really see how it's much baggage. It's straight out of Econ 101. Using very simple Econ 101 P=MR, MC, and ATC, we can show that the marginal firm earns zero profits while firms producing identical products at lower costs will earn profits.
As an aside, this example of Alaskan avocado farmers could also be used to illustrate the absurdity of a price-control intended to eliminate profits.
Suppose some politician said that it's not enough that the marginal firm earns zero profits. They want *all* firms to earn zero profits.
So they impose a price-control intended to set profits to zero.
Then Californian farmers have to charge their ATC and Alaskan farmers charge their ATC. Let's say it's $1 and $5 respectively. The market price of avocadoes must be $5 since the marginal Alaska farmer is earning zero profit.
So then the grocery store buys Californian avocadoes for $1 and resells them for $5. Instead of profits going to the farmer, they just go to the grocery store instead.
The politician then has to impose a price-control on the grocery store, forcing them to sell Californian avocadoes for $1 and Alaskan for $5. Now, the grocery store has to have two separate bins of identical avocadoes.
At this point, consumers will presumably buy the cheaper Californian avocadoes, letting the Alaskan avocadoes pile up until the Californian avocadoes are all gone. We know that total supply equals total demand, so Californian supply is not enough to meet total demand. So *some* consumers will have to buy Alaskan avocadoes. So we've created a first-come-first-served system where some consumers pay $1 and others pay $5 for the same avocadoes. It's not clear why that's any better or more fair than just letting Californian avocado farmers earn a profit.
And I'm not sure that that will achieve market equilibrium anyway. $5 was the market-clearing price that equated quantity-supplied with quantity-demanded. So the marginal consumer values avocadoes at $5. But what if the highest-value consumers all get there first and buy all the $1 avocadoes. Then the lower-value consumers - who value avocadoes at say $2 or $3 - get there and find only $5 avocadoes. So they don't buy any. Now we have a surplus.
Or maybe the grocery store tries to sell avocadoes for a price equal to the quantity-weighted average of $1 and $5. Let's say 2/3 of avocadoes are California and 1/3 are Alaskan. So the grocery store sells them all for $2.33, after buying 2/3 of their stock for $1 and 1/3 of their stock for $5. But then we get a shortage, because $5 was the market-clearing price that equated quantity-supplied and quantity-demanded. I'm not sure what the equilibrium is under this whole political scenario, but it's getting messy.
So I think that eliminating *all* profits from *all* stages of the supply-chain would require a perfectly price-discriminating politician.
Good article Brian. I get the feeling that we use so many models in intro micro that we confuse people as to the basic message. My only gripe with this piece is that you do use the word competitive without being too precise what you mean. You seem to mean something like, we reach the same intersection of the two curves, but I still want more explicit standards as to what would make a market "competitive". What kind of market would you call non-competitive? (Presumably something with high barriers to entry for firms). Anyway, good stuff. Big fan. When I teach micro I'll certainly be using some of these with students.
Thanks for the comment. I should have been more explicit.
When we draw supply and demand we are assuming people are price-takers. That’s same for a Walrasian equilibrium.
The complication is that we can always assume people are price-takers. But should they? So a competitive equilibrium is when everyone “should” be a price taker or would act as if a price taker, even if given the choice. So Bertrand is competitive since the equilibrium is the same as one where we forced everyone to be a price-taker.
Wanting to be a price-taker turns out to mean that each individual faces a perfectly elastic curve, at least locally. So unlike the standard monopolist with downward sloping demand, the perfect competitor can’t withhold goods to raise the price.
Does that make sense? We can always complicate things so that the competition is along a different dimension, such as competition for an exclusive contract compared to competition for the final marginal units sold.
Great piece, Brian.
Even though I agree with part regarding asymmetric information, I'm wondering if asymmetric information could come to trouble, especially in healthcare sector. Or should it be called "asymmetric chance" since patients have much less chance to prepare for health-related events, even more so in an accident. I believe you've heard this question a lot from your students but it'll reall help if you could enlighten me a bit. Thanks in advance!
It's an important question. Absolutely. We need to worry about it. The claim is not that any asymmetric information is fine. Some is. Some isn't. Some is compatible with the perfect competition model. Some is not. We need to dig in and be more specific in the case.
I agree with one of your conclusions, but I disagree with your proof. So I want to quibble a little.
You say that profits can exist under perfect competition because of the existence of producer surplus. But producer surplus is really profit over marginal/variable costs, ignoring fixed costs.
Producer surplus is the difference between the market price and the price the firm would have accepted. In the short-run, the firm will accept any price that is greater than its variable costs, even if it fails to cover its fixed costs. So the existence of producer surplus may merely indicate that firms are covering their variable costs, not that they are making profits (i.e. revenue exceeds total costs).
However, we can still prove your proposition in another way, by using the firm side rather than the market side. I.e., draw the perfectly competitive firm's graph, with horizontal P=MR, upward-sloping MC curve, and U-shaped ATC curve. The marginal firm will have P=ATC and make zero profits. But other firms producing an identical product at lower costs will earn the same P=MR but have a lower ATC, creating profits (P>ATC). Thus, the marginal firm makes zero profits, but all infra-marginal firms will make positive profits.
In my Econ 101 course, I use the example of growing avocadoes in Alaska. If the demand curve for avocadoes shifts right on the market side, then the P of avocadoes increases. All avocado farmers temporarily earn profits. In the long-run, supply increases, and the price falls back down until their profits disappear.* But here is where I make the next distinction: suppose that the way we increased the supply of avocadoes was by building greenhouses in Alaska. This clearly costs more than growing avocadoes using sunlight in California. So the Californian farmers will earn Ricardian rents. The marginal Alaskan farmer will earn zero profit because Alaskan avocado farmers will build greenhouses up until the point where the last greenhouse earns zero profit. But at that point, Californian avocado farmers are selling identical avocadoes with a lower ATC because of the free sun and warmth of California. So the marginal Alaskan avocado farmer earns zero profit, while the Californian avocado farmers all earn Ricardian rents.
I tell my students that the reason why profit disappears is that profit creates an incentive for competitive entry. If all firms are identical, then they all compete the price down to the cost of production. But if firms are not identical, then instead, firms enter until the last firm earns zero profit. Prior firms with lower costs will be earning profits. But the final firm that enters will earn zero profits, and no one else enters after that.
* (Aside: if it a constant-cost industry, then the long-run supply is perfectly elastic and price falls back down to its original level. If increasing-cost, then the long-run supply is upward-sloping - but more elastic than short-run supply - and the price falls back down only partway towards the original price. But either way, if all avocado farmers are identical, then in the end, all profits return to zero. Identical firms plus increasing-cost implies that costs for all firms increased identically. Perhaps increasing the supply of avocadoes caused an increase in the demand for fertilizer, so fertilizer prices increased equally for all farmers.)
I could have been more explicit about assuming fixed costs being zero. But I’m not sure why you want to import all the baggage of Viner’s curves and fixed vs. variable costs. Sure, that’s a story of how markets can get more competitive over time. Sure. But it’s not the definition of perfect competition. Even Mankiw makes this distinction.
I don't really see how it's much baggage. It's straight out of Econ 101. Using very simple Econ 101 P=MR, MC, and ATC, we can show that the marginal firm earns zero profits while firms producing identical products at lower costs will earn profits.
I agree with you that's a common feature of Econ 101. But I'm not making a claim about Econ 101. I'm making a claim about perfect competition.
As an aside, this example of Alaskan avocado farmers could also be used to illustrate the absurdity of a price-control intended to eliminate profits.
Suppose some politician said that it's not enough that the marginal firm earns zero profits. They want *all* firms to earn zero profits.
So they impose a price-control intended to set profits to zero.
Then Californian farmers have to charge their ATC and Alaskan farmers charge their ATC. Let's say it's $1 and $5 respectively. The market price of avocadoes must be $5 since the marginal Alaska farmer is earning zero profit.
So then the grocery store buys Californian avocadoes for $1 and resells them for $5. Instead of profits going to the farmer, they just go to the grocery store instead.
The politician then has to impose a price-control on the grocery store, forcing them to sell Californian avocadoes for $1 and Alaskan for $5. Now, the grocery store has to have two separate bins of identical avocadoes.
At this point, consumers will presumably buy the cheaper Californian avocadoes, letting the Alaskan avocadoes pile up until the Californian avocadoes are all gone. We know that total supply equals total demand, so Californian supply is not enough to meet total demand. So *some* consumers will have to buy Alaskan avocadoes. So we've created a first-come-first-served system where some consumers pay $1 and others pay $5 for the same avocadoes. It's not clear why that's any better or more fair than just letting Californian avocado farmers earn a profit.
And I'm not sure that that will achieve market equilibrium anyway. $5 was the market-clearing price that equated quantity-supplied with quantity-demanded. So the marginal consumer values avocadoes at $5. But what if the highest-value consumers all get there first and buy all the $1 avocadoes. Then the lower-value consumers - who value avocadoes at say $2 or $3 - get there and find only $5 avocadoes. So they don't buy any. Now we have a surplus.
Or maybe the grocery store tries to sell avocadoes for a price equal to the quantity-weighted average of $1 and $5. Let's say 2/3 of avocadoes are California and 1/3 are Alaskan. So the grocery store sells them all for $2.33, after buying 2/3 of their stock for $1 and 1/3 of their stock for $5. But then we get a shortage, because $5 was the market-clearing price that equated quantity-supplied and quantity-demanded. I'm not sure what the equilibrium is under this whole political scenario, but it's getting messy.
So I think that eliminating *all* profits from *all* stages of the supply-chain would require a perfectly price-discriminating politician.
_