« Defending Hayek against the Austrians | Main | Mark Carney and NGDPLPT »

Comments

Feed You can follow this conversation by subscribing to the comment feed for this post.

Shouldn't the random vertical stripes be denser towards the right? Isn't there still be some tendency for the marginal mutually beneficial trades to be made more often than the inframarginal ones?

Nitpick: the Okun gap is still the trapezoid. The double triangles would be the Okun gap under ideal policy that neutralizes the gap and leaves only the 'structural' losses, so to speak.

I really like these charts! I don't have anything productive to add but wanted to say that.

Off topic, but relevant:

Ever consider writing a layman's guide to, say, macroeconomics from Keynes to the present? You've a knack for actually explaining things that are totally unknown to the public and most policy makers: the difference between, say, Old Keynesian and New Keynesian models, and showing their implications in simple terms.

This is one of the things your blogging has been about, but there's a glaring gap in the printed literature. Even the educated public tend to think macro is a big debate between Keynes and Hayek (!), without really understanding either or anything else.

And hey, explaining the rest of macro would be a vehicle allowing you to get the monetary disequilibrium story out there as well.

I have to second Alex - all the power to the charts. And very interesting article by the way, I have never thought about it in this way. I need to process it.

PS: This also covers the old idea of yours - it is not only GDP that matters. There are trades on markets with previously produced goods and services that will be affected. People without jobs is only the most visible one. But empty - or not as well used houses on the market with used houses, or unsold things with sticky prices on Ebay are another source of DWL that will not be captured by official GDP statistics.

Thomas: maybe. Because the people who really really want the job or to make the sale join the queue earlier, or search harder, or polish their resume or apples more diligently, and so are more likely to get lucky. But then we get that blue rectangle in the first picture turning red too.

david: I don't think that's right. If we perfectly staibilses the economy, in the New Keynesian model, we would get the one red structural triangle in the first diagram (only it's due to monopolistically competitive firms, who have MR less than Price, setting Price at Pb, above MC, not to a tax).

Alex: thanks for saying that. It is productive, but in another way. It tells me what works, plus it makes me feel good.

Edmund: thanks. It would be good if I could do that, but I can't. When I'm blogging, you only see the things I can do (or think I can do). You don't see the things I can't do. And one of the things I can't do is concentrate for more than a few hours on completing any long project. (And spending my time reading and writing blogs has worsened that tendency.)

JV: Yes! I was thinking about linking to that old post I did where GDP was constant but we still had a recession, because people weren't trading houses or fruit trees or something. But I can't find it.

The fall in GDP is just a partial symptom of the costs of a recession. The costs of s recession are all the mutually beneficial trades that don't get made. The Okun Gap is a poor proxy.

JV, Nick, was the old post that you're thinking of this one: http://worthwhile.typepad.com/worthwhile_canadian_initi/2011/02/money.html? Though you've done a number of posts on apples.

B.t.w., I'm with Alex, like the pictures.

Frances: thanks! That wasn't the post (though it was related). But you have jogged my memory. This is the old post I was thinking of, on the Trade Cycle. I think it's the one JV was thinking of too?

Thinking again about Thomas' point. The blue rectangle in the first picture, when it turns red, can be thought of as a lot of vertical red stripes, each stripe representing the costs of queuing for one apple. If everybody who buys an apple has to queue, we get the solid red rectangle, plus the red triangle. If we ration by lottery, we get the random stripes in my last picture.

Hmmm. Those 2 red triangles in the New Keynesian model represent the average cost of business cycles over TWO periods (one period boom and one period recession). If you halve both the triangles horizontally, to get the average cost per ONE period, and then join the two halves together, you get ONE triangle.

(This now makes sense to me. My intuition was telling me it's one triangle, but the picture was telling me it's two triangles. It's really one triangle.)

Nick,

If the only thing that a distorting tax does is shift money from one type of good to the other (change in relative prices) then there should be marginal cost / marginal benefit curves that have a right hand side triangle. With left hand triangles there's a Net Welfare loss triangle, measuring the unexploited gains from trade.

What do you make of right hand side triangles where a combination of subsidies raises the price received for the seller and lowers the price paid for the buyer? How is this considered unexploited gains from trade when both buyer and seller are trading more goods?

Frank: those right hand side triangles are red too. People will trade too many goods if the goods are subsidised. If you trade goods where MB < MC, that's bad.

Those are trades that shouldn't happen.

(Normally though, that's not what governments do with the blue triangle. They subsidise goods where markets provide too little, creating a green trapezoid.)

Nick, I would argue that the main problem is wage stickiness, not monopolistic competition. Thus M-comp doesn't cause people to work a suboptimal number of hours.

The problem with my theory is that booms should feel "bad" as people work too hard. I explain the good feeling from booms by the fact that there are all sorts of governmental distortions that cause people to work less then they would if after-tax hourly income equalled MRP. So the boom pushes against all those labor market distortions, temporarily nullifying them. But just like a drug that takes away the pain from everyday life, the effects cannot last.

I suppose firms having monopsonistic power in the labor market should be added to my list.

Nick: Yes, "Why Y" was the post I had on my mind. That was the moment when I mentally stopped using "GDP" and replaced it by "newly-produced goods and services" as you did in that post. With definition of how new something has to be to be counted as newly produced. Which kind of reminds me, that it really is very arbitrary. If we are thinking of goods in Arrow-Debreu model, then the same good consumed in different time/place is basically a new good.

It is too cheap an accounting shortcut to say that everything ever produced has the value set at the prevailing market price during at the very moment it has been sold (except inventories that are priced by at the prevailing market price, which is kind of odd as we may think of investment and durable goods as a bunch of inventories to be sold again and again every accounting period). Imagine that someone would say that the value of Alaska with all its mineral riches is $7.2 million in 1868 dollars when it was purchased by USA and every barrel of oil or ounce of gold mined and sold is just a resell of that long-past sold piece of land with 0 impact on the GDP.


PS: I know that it is not related, but there was an article from Andy Harless (and some later on MM blogs including Sumner vs Krugman "What is saving" discussion) made me rethink the whole Saving vs Investment issue in a similare way - that arbitrary set accounting time artificialy divides things into artificial categories, like "newly produced stuff" and "old stuff" or "investment" and "consumption" . This is the article: http://blog.andyharless.com/2009/11/investment-makes-saving-possible.html

JV: "newly product stuff" vs "old stuff" is not artificial. The price elasticity of supply for "old stuff" is always and everywhere 0, which makes it very different from "newly produced stuff".

Ok, so why do we measure "newly" as "produced in current year"? Why not "produced and sold in last 5 minutes" or "produced and sold last century"? Why do we think that value of housing services that a house will produce for next decades are best evaluated at the moment of purchase and not at the moment the services are actually delivered?

Look, I understand why we do it like that and that it may be useful for many important things. But at the same time it is useful to always keep in mind that there are large gaps that this thinking is designed to omit. One of them is impact of monetary policy on DWL stemming from insufficient trade of goods not affecting "change in production of goods and services produced last year compared to *estimated value of the goods and services produced last year" ... *(the value is computed using deflator as measured by our estimate of the potential present value of qualitatively different product produced last year) That is all what I am saying.

PS: And by the way, I apologize. I don't want to hijack the discussion under this post. I am actually eager to see if there is anybody who has to say something on-topic.

Scott: I *think* that Nick means that it is monopolistic competition that makes us feel that we are richer in booms - because actually we are. That feeling stems from decreased DWL during booms. And actually you yourself said as much under this post: http://worthwhile.typepad.com/worthwhile_canadian_initi/2010/11/excess-supply-under-monopolistic-competition.html

"I believe this is why a monetary shock that pushes us to the left of the natural rate of output seems to be more destructive than a monetary shock that pushes us to the right of the natural rate, even though it really isn't more destructive in the long run."

JV:

Ok, so why do we measure "newly" as "produced in current year"? Why not "produced and sold in last 5 minutes" or "produced and sold last century"?

Because we can't actually measure instantaneous GDP, and integrating GDP over the previous year seems to give a halfway decent estimate. Why a year instead of 3 months, or 15? Social convention. But what we really really care about is instantaneous GDP, or instantaneous GDP integrated over some future period.

Why do we think that value of housing services that a house will produce for next decades are best evaluated at the moment of purchase and not at the moment the services are actually delivered?

Because they are best evaluated then; it would be much more difficult to accurately evaluate them when the services are delivered.

Look, I understand why we do it like that and that it may be useful for many important things. But at the same time it is useful to always keep in mind that there are large gaps that this thinking is designed to omit. One of them is impact of monetary policy on DWL stemming from insufficient trade of goods not affecting "change in production of goods and services produced last year compared to *estimated value of the goods and services produced last year" ... *(the value is computed using deflator as measured by our estimate of the potential present value of qualitatively different product produced last year) That is all what I am saying.

Yes, that's Nick's point with the fruit tree economy where I like apples and you like pears, but I'm stuck with the pear tree and you're stuck with the apple tree.

david: I don't think that's right. If we perfectly staibilses the economy, in the New Keynesian model, we would get the one red structural triangle in the first diagram (only it's due to monopolistically competitive firms, who have MR less than Price, setting Price at Pb, above MC, not to a tax).

True, and yet in the recessionary period, the recession will still feel particularly large. An Okun gap is not defined over long-run averaged GDP, it is defined in each period.

Nick,

"Frank: those right hand side triangles are red too. People will trade too many goods if the goods are subsidised. If you trade goods where MB < MC, that's bad."

Why is it bad? I thought we were talking economics here not religion or law. Where role does morality (good versus bad) play in economics? Ultimately when you shift from economic theory to economic policy good versus bad takes on a meaning of first, what goals are you trying to accomplish and second, what methods are you using to accomplish those goals. Policies that are conducive to those goals can be considered good while policies that are not can be considered bad.

And so I would agree with you, changes in relative prices caused by changes in tax policy (or monopolies, etc.) can be considered a bad thing, BUT only if they are not conducive to economic policy goals. And so to label an economic policy decision as bad, you first need to establish what economic goal that policy was meant to attain.

"People will trade too many goods if the goods are subsidised."

Is there any difference between subsidising the goods and subsidising the means of exchange to pay for those goods? And so, if a government feels that too many goods are being traded they should do what - suspend legal tender laws and force people to rely on a barter system or a disjointed set of local currencies hoping to slow the trade of goods? Legal tender laws are in fact a subsidy that the government provides that allow people to trade goods. Legal tender laws reduce the costs of engaging in the economic activity of trading goods for a common means of exchange. Without that common medium of exchange, people encounter transaction fees from swapping between this town / city's currency and that town / city's currency.

Scott: I think you need both: price (and/or wage) stickiness; plus some sort of market power. But I think you need monopoly power (rather than monopsony power) in the output and/or labour markets.

Frank:

Why is it bad? I thought we were talking economics here not religion or law. Where role does morality (good versus bad) play in economics?

In an economic context, normative claims will always be made relative to a sort of crude utilitarian framework. I'm sure you could challenge that as a proper basis for morality, but it will be way off-topic for most discussions.

Nick, Do you need price stickiness to get recessions, or the stylized facts we observe with actual recessions? I'd say no to the first question and yes to the second.

Alex,

"In an economic context, normative claims will always be made relative to a sort of crude utilitarian framework. I'm sure you could challenge that as a proper basis for morality, but it will be way off-topic for most discussions."

Good versus bad does not need to be based on a set of morals. It can be objective based as I previously described. And so when we say that a tax or subsidy is distortive, we are saying that assuming a utilitarian value system for goods, the goods will be traded where the cost will be greater than or less than the utilitarian value of the goods.

A couple problems with this. First, a value system of goods based solely on utilatarianism is incomplete and does not accurately reflect reality.

The second problem is when we start measuring the value of goods in terms of a government endorsed currency and taxes / subsidies are currency based. In that case the notion of a distortive tax / subsidy is hazy. The mere fact that a government provides a common means of exchange means that the government is providing a subsidy for the exchange of goods. And so right away the common means of exchange creates a right hand side triangle on the marginal benefit / marginal cost curve. The government then levies a tax to create a demand for this common currency. The taxation shrinks the size of the right hand triangle, but does it shrink it to the point that a left hand triangle is created?

Frank: that the government provides it doesn't mean there is a subsidy. If the government provides the exactly optimal quantity of currency, there is no welfare loss.

Alex,

When the government provides it in place of a mix of existing currencies or a barter system - yes there is a subsidy. If a government replaces one common means of exchange for another then there is no subsidy.

The quantity of currency is irrelevant as far as deciding whether a common currency is a subsidy. The subsidy is in reducing friction losses that occur when a mix of existing currencies must be exchanged one for another to engage in commerce on a large scale.

Frank, do you have a coherent definition of "subsidy"? Are you sure it means what economists mean? Do you realize that a "subsidy" does not always cause a welfare loss?

Alex,

http://en.wikipedia.org/wiki/Subsidy
"A subsidy is assistance paid to a business or economic sector or producers."

The assistance is provided by the government in the form of creating a common currency that all economic sectors can use. It is not a payment per se, instead it provides a reduction in transaction costs.

"Do you realize that a subsidy does not always cause a welfare loss?"
Yes I am very well aware of it. Just because the government offers a subsidy does not mean that the recipient has to accept that subsidy. And so while the subsidy exists, it may not be taken advantage of (no welfare loss is realized).

Alex,

What I am getting at is this. Because a government can create a common currency and because it can create oodles of that currency on a whim, that alone will push the marginal cost of goods above the marginal benefit of goods UNLESS a tax policy exists to pull the marginal cost back in line with the marginal benefit.

Why does a government borrow - to create a supply of it's currency
Why does a government tax - to create a demand for it's currency

"If the government provides the exactly optimal quantity of currency, there is no welfare loss."

It is not about quantity. It is about matching supply and demand. There is not a singular optimum quantity that matches supply of money with demand for money because of interest rates (cost of money). There is a supply and demand curve for money just like any other good. Find the intersection of the supply curve for money (remembering that government borrowing adds to the supply) and demand curve for money (remembering that taxes add to the money demand) and you find an interest rate and a quantity of money. Compare that to the intersection of the supply and demand curve for goods.

Frank:

What I am getting at is this. Because a government can create a common currency and because it can create oodles of that currency on a whim, that alone will push the marginal cost of goods above the marginal benefit of goods UNLESS a tax policy exists to pull the marginal cost back in line with the marginal benefit.

Um. No? Why are people engaging in these transactions if MC > MR?

It is not about quantity. It is about matching supply and demand. There is not a singular optimum quantity that matches supply of money with demand for money because of interest rates (cost of money). There is a supply and demand curve for money just like any other good.

Actually, there are N supply and demand curves for money in a monetary exchange economy, where N is the number of different goods/services/assets that can be bought or sold for money. But this doesn't mean it isn't about quantity; supply and demand curves are how we figure out quantity (and price).

Alex,

"Um. No? Why are people engaging in these transactions if MC > MR"

First I thought we are looking at marginal cost versus marginal benefit (not marginal revenue). Either way, with a government the creating money (borrowing it into existence), cost constraints go out the window. A government does not face bankruptcy or solvency constraints like an individual or firm does. In the case of cost versus revenue, with a sufficient subsidy the firm can operate at a marginal loss while staying solvent or even profitable provided that the subsidy is perpetuous. Why would the firm engage in transactions if MC > MR - if it can and remain solvent or even turn a profit, why wouldn't it?

"Actually, there are N supply and demand curves for money in a monetary exchange economy, where N is the number of different goods/services/assets that can be bought or sold for money. But this doesn't mean it isn't about quantity; supply and demand curves are how we figure out quantity (and price)."

Money has its own supply / demand curve not totally dependent on goods because of liquidity preference. People don't spend money the instant that they get it nor are they able to re-lend that money the instant they deposit it into a bank. But you are getting my point - it isn't just about quantity. A government does not try to provide an optimal quantity of currency. Through monetary policy it tries to regulate the cost of money. Through fiscal policy it tries to regulate the supply and demand for money.

Frank:

Either way, with a government the creating money (borrowing it into existence), cost constraints go out the window. A government does not face bankruptcy or solvency constraints like an individual or firm does. In the case of cost versus revenue, with a sufficient subsidy the firm can operate at a marginal loss while staying solvent or even profitable provided that the subsidy is perpetuous. Why would the firm engage in transactions if MC > MR - if it can and remain solvent or even turn a profit, why wouldn't it?

You still haven't explained where this mysterious subsidy comes into play. Where is the transaction happening where marginal cost is less than marginal benefit? Why is that transaction happening?

Money has its own supply / demand curve not totally dependent on goods because of liquidity preference.

No. Supply / demand curves are always about the exchange of one good/service/asset for another good/service/asset. You can't have a supply/demand curve for "money", only one for "money in exchange for X". X can be bonds, X can be apples, X can be haircuts, but X has to be something.

Alex,

"No. Supply / demand curves are always about the exchange of one good/service/asset for another good/service/asset. You can't have a supply/demand curve for "money", only one for money in exchange for X. X can be bonds, X can be apples, X can be haircuts, but X has to be something."

No. You can have a debt without a bond (credit cards). The reason that bonds exist is first to maintain a legal record that a loan was agreed upon (for long term loans) and second so that they can be sold by the issuing agent (lender) to a third party. And so, yes you can have a supply / demand curve for money at an interest rate without bonds. Y axis is price of money (interest rate). X axis is quantity of money. Demand curve falls with price (interest rate), supply curve rises with price (interest rate). Imagine a bank that makes loans and retains all of the loans that it makes.

"You still haven't explained where this mysterious subsidy comes into play. Where is the transaction happening where marginal cost is less than marginal benefit? Why is that transaction happening?"

Okay so we are back to marginal cost versus marginal benefit. Why would the buyer engage in additional transactions if the marginal cost is less than the marginal benefit? Vanity would be one case. Charity would be another. Faced with a declining marginal cost for the same benefit a person could either decide to redefine what level of benefits they wish to obtain (vanity) or they could chose to extend those benefits to friends, family, neighbors, etc (charity).

To create a marginal cost / marginal benefit graph we must lock down the value system we are using to measure cost and benefit (for instance you mentioned utilitarianism). When that value system changes (vanity and charity) we can either say that the marginal cost is below the marginal benefit or we can create a new graph and redefine our value system.


With "why Y", your're crossing over into Keen territory. For someone so critical of him, you share remarkably similar opinions.

"There's something really wrong with the way we do short run macroeconomics. We focus all our attention on the output of newly-produced goods and services. That's what we call "Y". We talk about Aggregate Demand and Aggregate Supply, and what we mean by AD and AS is the demand and supply of those same newly-produced goods and services.

Keynesians then go on to divide Y into C+I+G, and C+S+T. Monetarists talk about MV=PY. Both agree that a recession is a fall in Y, caused by a drop in demand for Y.

But a moment's reflection tells you this is wrong. It's not just new stuff that is harder to sell in a recession; it's old stuff too. New cars and old cars. New houses and old houses. New paintings and old paintings. New furniture and antique furniture. New machine tools and old machine tools. New land and old land."

Keen's "mysterious" equation is an attempt to answer your question:

"I have no idea how I would modify Keynesian macroeconomics to handle an economy like that."

He takes Y as income and adds in net credit(what's available to spend) which he equates to production + realized changes in existing assets (what was bought). This is "a monetary disequilibrium approach to macro."


Try starting with his short talk:


http://www.debtdeflation.com/blogs/2012/12/01/the-debt-issue-in-neoclassical-economics/

BTW there's nothing wrong with the first part of his marginal cost paper. It is an attack on the assumption of atomism. It depends on the off-diagonal terms of the differential being 0. This is obviously counterfactual. He is objecting to its being taught in undergraduate economics, since it is wrong with regard both to mathematics and the real economhy.

The second half of the paper is much more interesting.

Frank

No. You can have a debt without a bond (credit cards). The reason that bonds exist is first to maintain a legal record that a loan was agreed upon (for long term loans) and second so that they can be sold by the issuing agent (lender) to a third party. And so, yes you can have a supply / demand curve for money at an interest rate without bonds. Y axis is price of money (interest rate). X axis is quantity of money. Demand curve falls with price (interest rate), supply curve rises with price (interest rate). Imagine a bank that makes loans and retains all of the loans that it makes.

Money due at some point in the future (colloquially, a "bond" regardless of the details) is not the same asset as money today. The idea of supply and demand curves for money alone is an absurdity, because no one needs to exchange money for money.

Why would the buyer engage in additional transactions if the marginal cost is less than the marginal benefit? Vanity would be one case. Charity would be another.

1) What the heck does this have to do with money, let alone government money?
2) Why do you think neither of those counts as "benefit"?

To create a marginal cost / marginal benefit graph we must lock down the value system we are using to measure cost and benefit (for instance you mentioned utilitarianism).

No, we really don't. We just start with the general supposition that the value of X to a person is equal to the value of what he would be willing to offer in exchange, and add details to the model as necessary for whatever analysis we're performing.

You really need to take Nick's advice and read a first-year econ textbook. You keep building models of the world that have way too much detail, and they suffer for the fact that you don't seem to have any experience building *simple* models.

Alex,

"Money due at some point in the future (colloquially, a "bond" regardless of the details) is not the same asset as money today. The idea of supply and demand curves for money alone is an absurdity, because no one needs to exchange money for money."

Is there a supply of money?
Is there a demand for money?
Is there a cost of money?

"You keep building models of the world that have way too much detail, and they suffer for the fact that you don't seem to have any experience building *simple* models."

Simple model, cost of money (interest rate) on Y axis, quantity of money on X axis, supply curve slopes up from left to right, demand curve slopes down from left to right.

See forex markets for money / money exchanges.

"You still haven't explained where this mysterious subsidy comes into play. Where is the transaction happening where marginal cost is less than marginal benefit? Why is that transaction happening?"

Can we agree that a common shared currency reduces marginal costs? Can we agree that we are using a utilitarian value system for the marginal benefit of goods - meaning we are using a value system that does not include cost as part of the value? And so if a government would come along and unite 10 currencies with one currency, marginal costs of goods drop while the marginal benefit of goods remains the same.

Was that not the promise of the Euro?

Alex Godofsky: 'The price elasticity of supply for "old stuff" is always and everywhere 0, which makes it very different from "newly produced stuff".'

Hmm... if I bid double the market value for a house that is "not for sale," it is very likely going to become for sale. The talent of a retired entertainer could be considered "old stuff": the human capital was produced long ago. A big enough paycheck will lure her out of retirement. When the Hunt brothers tried to corner the silver market and drove the price way up, people dug their silverware out of the attic and sold it. Why don't these facts mean the elasticity of this old stuff is non-zero?

Gene and Alex: there is more than one way to think about the supply curve of old stuff:

1. The gross stock of vintage cars owned by someone. Perfectly inelastic maybe, but even then if the price falls people won't maintain them, so some will collapse into rust and the stock will fall.

2. The flow of those vintage cars being put on the market this year by their current owners. That will slope up.

Sir.

Just read this post.

Quality.

The comments to this entry are closed.

Search this site

  • Google

    WWW
    worthwhile.typepad.com
Blog powered by Typepad