Everything is not IN equilibrium, I would argue. "Everything is an equilibrium" is a worldview to understand social phenomena, but obviously never a reality. The market is moved by gaps in information that need to be detected. Assuming away the knowledge problem does not seem to be the way forward to understand market phenomena. (Nothing against equilibrium analysis). Problem is to assume away the informational issues that are present in the market economy. Assuming there is a knowledge issue, it seems obvious that the market is not instantaneously adjusted, that there are delays, errors, frictions --assumed away in Eco 101 and up.
I like Baumol's comment about economics as a "Hamlet without a Prince", a world where entrepreneurs do not have much to do!
Thanks for the piece Brian, I find it to be thoughtful and well-considered. I have two questions though, if you'll forgive my being a bit blunt: 1) how is this different from the Post-Keynesian price theory that many of us in the heterodox world subscribe to, and 2) how do you square this with a lot of the other material I see on this blog, which seems to rely on prices clearing markets?
Hey Sam. Thanks. On 1, I don't claim to be an expert of Post-Keynesian price theory. I've read Lee's book but cannot claim to understand what it was saying. I think you'd be the better one to explain. What do you see as the differences? But you may like the discussion of administered prices in the Carlton piece I linked to. On 2, I'd say it depends on the question at hand. For example, if I want to explain that output prices will rise when input costs rise, that doesn't really require me to worry too much about what happens to waiting times. To the extent other models give the same prediction, that's great. You can stick with the easiest model model to work with. But if I want to think about timing of price changes in some shorter window, then pure price models seem to miss the mark. Yeah. Absolutely. Of course, it's rarely an either or between prices and waiting.
On 1, I think a lot of what you've said here fits well with the general Post-Keynesian conception of the firm, which sees firms as long-lived "going concerns," in an evolutionary environment (you'll like this from Alchian), who face uncertainty and must manage various goals, first and foremoest of which is survival. They are also social enterprises, i.e. sites of cooperation and conflict, with various stakeholders who interact in various ways, and who abide in a world of laws and social norms. In short, they have agency and do not just mechanically maximize short-term profit based on marginal cost. This view emphasizes non-price competition, flat or falling cost curves, and quantity-adjustment rather than price adjustment. If you don't like Lee, another good reference would be chapter 3 out of this book: https://www.elgaronline.com/monobook/book/9781839109621/9781839109621.xml. (Particularly sections 3.5, 3.6, and 3.7.4).
On 2, thanks for the reflections. I'm going to disagree methodologically though with this part: "To the extent other models give the same prediction, that's great. You can stick with the easiest model model to work with." For example, suppose that you are wondering whether an increase in input price will lead to an increase in output price. The standard perfect competition model says the answer is yes. But I can conceive of an even simpler model which says the same thing: "whenever anything changes, output price increases." Taking the your statement literally, we should go with the 2nd model because it's simpler, but plainly there's something deeply flawed about that model and we should not be using it as our guide to anything. Contra Friedman, I think "giving the right prediction" isn't enough, there has to be some correspondence between elements in the model and elements in the real world, which gives us reason to believe that the model will be able to make good predictions out-of-sample, or else we can't trust it.
I agree we want more than prediction. I've written a ton about that. But just on the prediction point, I think we are thinking about prediction differently. I don't need a model to predict whether the sun will rise. I need a model to help sort out when something will happen vs. when it will not. It needs to have the possibility of type 1 and type 2 errors. I understand if that's not the only way that people use the word prediction but that's what I meant.
Thank you for your article. I think the source of the challenges that these models represent is the fact that they are equilibrium models. What we usually call an explanation of a price adjustment is in fact simple comparative statics around the equilibrium price--why the price should move up or down to the equilibrium price. In equilibrium it is not possible (not even necessary) to explain the process by which that equilibrium was formed. Modern economics does not have a satisfactory explanation of how prices, as well as non-price variables, behave outside the equilibrium state. In the equilibrium state everything is known or assumed to be known. The knowledge assumption (i.e. that everything that needs to be known is somehow known), makes it very difficult to explain price (and non-price) adjustments. I guess the only way out is to assume that you start with limited knowledge on the part of market participants, knowledge that tends to be revealed as the competitive process unfolds (the leading scholar on this topic is Israel M. Kirzner, Competition & Entrepreneurship.
I take a different approach to equilibrium. Everything is an equilibrium. If monetary price doesn't clear the market, something does. That must occur. The question is whether we are explicit about what is clearing the market or whether we simply say "disequilibrium" meaning we don't know. It's fine to say we don't know. Often we don't. But it's not really a way forward for how to understand the world.
I understand not everything is allocated by willingness to pay. But this isn't exactly the equitable way to do it either, favoring workers like me who are flexible. An easy fix would be random allocation for anyone who signs up in a week. That would decrease wasted waiting time and avoid allocating by price (if you want that).
I like your proposed idea here -- especially if the allocation was timely enough to allow the unfortunates sufficient time to make other arrangements. I was just noting that your original reaction: "compete on ... willingness to pay." is the reaction of someone who feels comfortable that their wealth would allow them to win the "pay" competition; it is unlikely that a lower earning individual would jump right to this solution.
Everything is not IN equilibrium, I would argue. "Everything is an equilibrium" is a worldview to understand social phenomena, but obviously never a reality. The market is moved by gaps in information that need to be detected. Assuming away the knowledge problem does not seem to be the way forward to understand market phenomena. (Nothing against equilibrium analysis). Problem is to assume away the informational issues that are present in the market economy. Assuming there is a knowledge issue, it seems obvious that the market is not instantaneously adjusted, that there are delays, errors, frictions --assumed away in Eco 101 and up.
I like Baumol's comment about economics as a "Hamlet without a Prince", a world where entrepreneurs do not have much to do!
Thanks for the piece Brian, I find it to be thoughtful and well-considered. I have two questions though, if you'll forgive my being a bit blunt: 1) how is this different from the Post-Keynesian price theory that many of us in the heterodox world subscribe to, and 2) how do you square this with a lot of the other material I see on this blog, which seems to rely on prices clearing markets?
Hey Sam. Thanks. On 1, I don't claim to be an expert of Post-Keynesian price theory. I've read Lee's book but cannot claim to understand what it was saying. I think you'd be the better one to explain. What do you see as the differences? But you may like the discussion of administered prices in the Carlton piece I linked to. On 2, I'd say it depends on the question at hand. For example, if I want to explain that output prices will rise when input costs rise, that doesn't really require me to worry too much about what happens to waiting times. To the extent other models give the same prediction, that's great. You can stick with the easiest model model to work with. But if I want to think about timing of price changes in some shorter window, then pure price models seem to miss the mark. Yeah. Absolutely. Of course, it's rarely an either or between prices and waiting.
Thanks for the response Brian!
On 1, I think a lot of what you've said here fits well with the general Post-Keynesian conception of the firm, which sees firms as long-lived "going concerns," in an evolutionary environment (you'll like this from Alchian), who face uncertainty and must manage various goals, first and foremoest of which is survival. They are also social enterprises, i.e. sites of cooperation and conflict, with various stakeholders who interact in various ways, and who abide in a world of laws and social norms. In short, they have agency and do not just mechanically maximize short-term profit based on marginal cost. This view emphasizes non-price competition, flat or falling cost curves, and quantity-adjustment rather than price adjustment. If you don't like Lee, another good reference would be chapter 3 out of this book: https://www.elgaronline.com/monobook/book/9781839109621/9781839109621.xml. (Particularly sections 3.5, 3.6, and 3.7.4).
On 2, thanks for the reflections. I'm going to disagree methodologically though with this part: "To the extent other models give the same prediction, that's great. You can stick with the easiest model model to work with." For example, suppose that you are wondering whether an increase in input price will lead to an increase in output price. The standard perfect competition model says the answer is yes. But I can conceive of an even simpler model which says the same thing: "whenever anything changes, output price increases." Taking the your statement literally, we should go with the 2nd model because it's simpler, but plainly there's something deeply flawed about that model and we should not be using it as our guide to anything. Contra Friedman, I think "giving the right prediction" isn't enough, there has to be some correspondence between elements in the model and elements in the real world, which gives us reason to believe that the model will be able to make good predictions out-of-sample, or else we can't trust it.
I agree we want more than prediction. I've written a ton about that. But just on the prediction point, I think we are thinking about prediction differently. I don't need a model to predict whether the sun will rise. I need a model to help sort out when something will happen vs. when it will not. It needs to have the possibility of type 1 and type 2 errors. I understand if that's not the only way that people use the word prediction but that's what I meant.
Thank you for your article. I think the source of the challenges that these models represent is the fact that they are equilibrium models. What we usually call an explanation of a price adjustment is in fact simple comparative statics around the equilibrium price--why the price should move up or down to the equilibrium price. In equilibrium it is not possible (not even necessary) to explain the process by which that equilibrium was formed. Modern economics does not have a satisfactory explanation of how prices, as well as non-price variables, behave outside the equilibrium state. In the equilibrium state everything is known or assumed to be known. The knowledge assumption (i.e. that everything that needs to be known is somehow known), makes it very difficult to explain price (and non-price) adjustments. I guess the only way out is to assume that you start with limited knowledge on the part of market participants, knowledge that tends to be revealed as the competitive process unfolds (the leading scholar on this topic is Israel M. Kirzner, Competition & Entrepreneurship.
I take a different approach to equilibrium. Everything is an equilibrium. If monetary price doesn't clear the market, something does. That must occur. The question is whether we are explicit about what is clearing the market or whether we simply say "disequilibrium" meaning we don't know. It's fine to say we don't know. Often we don't. But it's not really a way forward for how to understand the world.
"The process seemed inefficient and even unfair in making parents compete on speed and flexibility rather than willingness to pay."
If I interpret you correctly, you object to parents competing on a basis other than personal wealth. Amazing? Yes. Surprising. No.
I understand not everything is allocated by willingness to pay. But this isn't exactly the equitable way to do it either, favoring workers like me who are flexible. An easy fix would be random allocation for anyone who signs up in a week. That would decrease wasted waiting time and avoid allocating by price (if you want that).
I like your proposed idea here -- especially if the allocation was timely enough to allow the unfortunates sufficient time to make other arrangements. I was just noting that your original reaction: "compete on ... willingness to pay." is the reaction of someone who feels comfortable that their wealth would allow them to win the "pay" competition; it is unlikely that a lower earning individual would jump right to this solution.