In an unusually critical and mainly on-target criticism of one commenter’s critiques of Amazon, Tyler Cowen writes:
First, monopsony and monopoly tend to have contrasting or opposite effects. To the extent Amazon is a monopsony, that leads to higher output and lower prices.
The second sentence in the above quote is false. I don’t believe that Amazon is a monopsony but if it were, output would be lower and prices would be higher. That is a well-known result in economics. Indeed, one of my criticisms of David Card’s and Alan Krueger’s book, Myth and Measurement: The New Economics of the Minimum Wage, is that their argument that the minimum wage increase in New Jersey reduced monopsony power in the fast-food labor market is inconsistent with their finding that prices rose more in New Jersey, where the minimum wage increased, than in neighboring Pennsylvania, where it didn’t.
READER COMMENTS
blink
Sep 19 2018 at 8:38pm
Rather than disagreeing, it seems like the two of you are talking about different monopsonies. Cowen, I believe, considers the possibility that Amazon is a monopsony buyer of good for resale. By contrast, I understand you as explaining the effects of Amazon being a monopsony in a factor (labor) market. There is no reason either of you need be mistaken.
Pierre Lemieux
Sep 19 2018 at 9:13pm
Blink: Aren’t inventories a factor for a retailer as much as labor (to man the check-out, shelf the product, etc.) is?
David Henderson
Sep 20 2018 at 12:14am
Whether the item being monopsonized is wholesale goods or labor, the analysis of monopsony is the same. So Tyler is definitely mistaken.
Iskander
Sep 20 2018 at 12:34am
I think Tyler’s point holds when Amazon has monopsony power for labour but sells in a competitive market.
One employer in certain areas, immobile labour and tradeable goods?
David Henderson
Sep 20 2018 at 9:32am
I’m pretty sure that’s false. The reason: if Amazon is a monopsonist in the labor market, it hires fewer laborers than otherwise. That means lower output.
Pierre Lemieux
Sep 21 2018 at 12:42pm
David: It seems to me that a monopsony in the factor market means lower output only if there is just one factor (say, one sort of labor, or “captive” labor in one isolated geographic place). But then, if there is only one factor, the monopsonic firm is also a monopoly in the product market.
David Henderson
Sep 21 2018 at 1:34pm
Pierre,
I’m pretty sure that’s wrong. Think of the standard graph we draw for monopsony. The key is an upward-sloping supply curve of the factor that is monopsonized. There is no assumption that it is the only factor of production. It’s true that the firm will substitute into capital and so the reduction in output will be less than if it didn’t. But there will be a reduction in output.
Pierre Lemieux
Sep 21 2018 at 5:39pm
David: Clearly, there will be a reduction in the quantity demanded and price of the factor. But when this is transposed to the product market (assuming indeed that there is more than one factor), the monopolistic firm will see a downward shift in its marginal cost curve. How could this not lead to a lower price and higher quantity demanded on the product market? Isn’t this what Tyler is arguing? The way to deny this conclusion, it seems to me, is to argue that the monopsonist is a perfectly competitive firm on the product market (if that is possible), but then how could there be any impact on price and quantity (by definition of perfect competition)? I don’t see how intermediate cases (between monopoly and perfect competition on the product market) could lead to an increase in the market price or a reduction in market supply.
Robert Simmons
Sep 21 2018 at 3:17pm
David, can you make a post explaining monopsony’s effects?
David Henderson
Sep 22 2018 at 10:54am
Reply to Pierre and Robert Simmons.
I’m doing it this way so that others who might wonder will see it more easily.
Robert,
I will but it might or not be this weekend.
Pierre,
Ah. Now I see the problem. You’re assuming that marginal cost is lower with monopsony. In fact, it’s higher. That, in fact, is the monopsony problem. As noted above, I will try to get to this for Robert but meanwhile draw a few graphs and I think you’ll see it.
Pierre Lemieux
Sep 22 2018 at 12:22pm
David: Right! What you have to show in your demonstration, then, is that a firm that pays less for one input and the same price for all other inputs has a higher marginal cost curve on the product market.
Dan Culley
Sep 23 2018 at 9:12am
I think there’s a legitimate argument when one is talking about a platform that output may not be restricted on what would be the platform’s intensive margin, sales of an individual book, and instead wildw be restricted on the extensive margin, the number of titles sold. But I don’t see a way around output going down somewhere.
I don’t understand Pierre’s argument. Certainly the degree of marginal cost increase (and thus output distortion) depends on the substitutability of other production factors. But marginal cost never goes down from exercising monopsony power, as that’s the whole definition, that you restrict your input purchases to push down the price. Purchasing more of the input undoes some of the leverage you got by restricting purchases, and so the marginal cost is the sum of what you have to pay for the incremental unit, plus the additional amount you have to pay on the existing units you were buying because you lost some of your leverage. (Exactly the same reason that marginal revenue is always less than the revenue in the next unit sold for a firm with any market power.) If there were another factor that were perfectly substitutable, then there is no increase in marginal cost (since you just purchase the other factor), but it can’t be a cheaper factor otherwise you would have already switched to it in the first place.
Plus, Tyler was taking about distributing books where the monopsonized input was the books. What exactly is the other factor that you substitute to? Maybe he meant to other titles, but as above that to me just moves the distortion to another margin.
I suppose one could say that Amazon is actually exercising countervailing bargaining power against authors that have market power. And that in all these individual negotiations the parties choose a more competitive level of output and Amazon just takes more of the surplus. But pretty hard to believe that “authors” overall have meaningful market power, or that all those negotiations are landing on the right output level.
So I’m still not seeing it, unless what Tyler is really saying is that they both have monopsony power and benefit from economies of scale, and the latter are reducing marginal costs faster than the former are increasing them. But that of course means its costs are lower despite, not because of, monopsony power.
Comments are closed.