« Sovereign insolvency and illiquidity | Main | Are the rich capitalist? Are capitalists rich? »

TrackBack

TrackBack URL for this entry:
http://www.typepad.com/services/trackback/6a00d83451688169e20148c67bf172970c

Listed below are links to weblogs that reference Why blogging is hard:

Comments

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

You guys are using imaginary math constructs like this within democratic societies as grounds for multi-trillion dollar bailouts of insolvent mega-banks and corporations and other institutions controlled by the super rich -- so you'd better learn how to explain it to the people.

I wouldn't feel bad Nick, it was clear in an exchange between Winterspeak and Rebel Economist (that appears in the comments before you show up) that neither of those two have a handle on what inflation actually is or what it implies.

I've no doubt the rest of the commenters on the thread would be even worse than Rebel or Winterspeak in their understanding (excluding you of course).

So the only thing to do is laugh.

It is sort of funny, all round. But those guys genuinely want to understand, and they're not daft either, because they all know stuff that I don't know. As Greg says, it's our job to try to explain what we think we know, and why we believe it. But God it gets hard sometimes. And the temptation to really cheat, to cook up some story that sounds good but isn't really, is strong. But yes, let's have a laugh at our predicament too.

For example: I cheated by:

translating "increase in expected inflation" into "increase in expected future price level"

assuming the nominal interest rate stayed constant

ignoring open economy

using perfect competition, which I don't believe in

assuming an across the board increase in the expected future price level

assuming perfectly flexible prices, which I don't believe in

ignoring all distribution effects on nominal contracts

ignoring the seigniorage profits.

Many of those are necessary for the thought experiment to have no real effects, which is what I assumed.

Well, yeah, that's probaly true. Certainly both Winterspeak and Rebel are fairly serious.

On the other hand, I've seen the market player comment on several blogs and the impression he gives is of someone who is daft and is not serious about understanding anything.

BTW: I'm pretty sure the russian guy's handle translates as "market player" and not "short market".

Nick: "You have 2 hours to answer this question."

Almost. How about:

"This is a take-home exam. You may begin writing at 11:00 p.m. Please hand in your completed paper by 8:00 a.m."

and

"Examiners who make inappropriate or personal remarks may be deleted."

Indeed, and of course there's the all important part that the "examiners" in blog comments don't have any actual power over you, certainly they can't change the course of the rest of your life.

Phd examiners can change the course of the rest of your life.

"But those guys genuinely want to understand, and they're not daft either..."

I just read through this post: http://www.winterspeak.com/2010/11/krugman-is-part-of-problem.html

and this post: http://www.winterspeak.com/2010/11/qe2-is-about-rates-not-prices.html.

As far as I can tell Winterspeak knows *less* than nothing about macro and is not genuinely trying to understand.

"I've no doubt the rest of the commenters on the thread would be even worse than Rebel or Winterspeak in their understanding (excluding you of course)."
__

Are you including yourself in that group Adam P?

I wasn't a commenter on the thread.

Frances: Ah! You saw what time I was commenting, rather than being sound asleep like any sensible person.

Steady guys. Winterspeak and commenters were civil to me, despite my slight loss of patience. Let's maintain that.

Nick, actually the scary things is: I didn't.

Blogging is actually a variation on the old chestnut: "Set your own question and answer it. You will be graded on both the quality of the question and the quality of the answer."

Alternatively: "Set up your own straw man and knock him down. You will graded on the size of the debris field."

For what it's worth, I have zero formal education in economics, and I found your basic explanation convincing and clear. (producers want to sell things for more money in the future, while buyers want to buy things for less money right now. that seems pretty intuitive).

I don't really understand why some of the commenters at that blog are so against the idea of inflation expectations. I can't imagine what kind of political or theoretical commitments would lead someone there. Is it just antipathy towards economists?

"You guys are using imaginary math constructs like this within democratic societies as grounds for multi-trillion dollar bailouts of insolvent mega-banks and corporations and other institutions controlled by the super rich -- so you'd better learn how to explain it to the people."

I don't know, Greg. Math is the language of quantities and change. Personally, I think "the people" could use a better grasp on it.

Stephen Bank: Thanks! And that makes my attempt worthwhile. Plus, you got the main message. Plus, you said "*want* to sell things", and "*want* to buy things", which shows you understand the difference between quantity supplied, quantity demanded, and quantity actually traded.

My internet experience tells me that you've come up against some economic doctrinaires. They're a really tenacious bunch and have their own body of knowledge. The discussion thread was hard to understand, so I don't think I can I identify them well. They sounded anti monetary policy. I've argued with someone like that before. One day they're advocating deflation as a way to encourage demand. You can argue against them then, but the next day they come out touting the gold standard and Max Keiser. It's like peeling an onion into a parallel universe.

I think the best way to deal with them is to adopt their way of thinking and debate the best you can using logic and technicalities. It's also a good idea to keep your messages short, since the internet is mostly edutainment anyway.

Nick, seriously, here's a quote of Winterspeak in the comment section of this *http://www.winterspeak.com/2010/11/qe2-is-about-rates-not-prices.html* post.

Winterspeak: When people expect inflation they switch from the financial asset they expect to depreciate to another financial asset that they hope will not depreciate.

This is *not* consumption.

The (garbage) macro models tie inflation expectations to velocity via the consumption channel (ie. households will move consumption forward).

You don't see a problem here?

The issue is not whether or not Winterspeak is smart or has some decent economic ideas. The problem is that Winterspeak has concluded a class of macro models is garbage without any understanding of what the model actualy says, nor is their any effort to undertand.

Just unquestioning dismissal of macro theory with a simple minded critique.

Adam: Yes, but. Suppose I argued like this:

"Well, if I thought inflation was coming, I would sell my bonds and buy a farm. And if a lot of people thought like me, the price of bonds would go down and the price of farms would go up. I'm not any richer if I think inflation is coming. I may be poorer, since my bonds are worth less. And since I'm not any richer, I'm not going to go out and consume more. I would probably consume less, since I'm poorer. So, since consumption won't go up, and might even go down, I can't see any reason to think that the CPI would go up today, just because I, and people like me, think it will go up tomorrow."

It is very hard to start with reasoning like that, then add in intertemporal substitution in consumption, and investment demand. And add in all the substitution effects from one asset to another. And, on top of that, to add in the implied monetary policy reaction function that is at the back of our minds (because, for example, we get very different explanations if we assume a fixed nominal money supply or a fixed nominal interest rate, since we get very different LM curves in each case).

Macro is hard. And we make it easy for ourselves by using simple models with a very limited class of goods and assets. What we do looks a bit like cheating, for anyone who cannot or will not abstract from the world in all its complexity.

And Winterspeak was spot on several months back in saying the Eurozone does not have a lender of last resort.

"And look, even if there is an increase in consumption, monopolistically competitive firms, facing horizontal MC curves, will just expand output in line with the increased demand. They won't raise prices!"

When I was working on my own macro model with monopolistic comp firms, before Blanchard and Kiyotaki in 87, it was very hard for me to get my head straight on why, at the macro level, an increase in demand would cause firms to raise prices, even when it might not do so at the micro level (with horizontal MC curves and isoelastic demand curves).

Sebastien has got it right. These discussions always circle back to the same themes: the glory of gold, the pernicious unfairness of inflation, the virtue of saving, the foolishness of borrowing. Oh, and "scrap the Fed", of course.

It's nice of you to answer questions that seem honestly posed, but don't be baited. Humour and a light touch are usually best. One of my favourites: Barry Ritholtz, fed up with inflationistas pricing stuff in gold terms, posted a graph of gold in units of silver (priced in silver, gold is going to zero), and proclaimed: "Quantitative Mining is debasing gold!"

Explaining things is hard. Persuading folks with a different understanding of things is even harder.


Let me recommend this article by Larry Wright on the issues involved,"Argument and Deliberation: a Plea for Understanding," The Journal of Philosophy, Vol. 92, No. 11, 1995, 565-85.

Formalism fails as a strategy for compelling assent -- it often doesn't even represent the phenomena in a valid way (e.g. math inspired formalism as a strategy for representing language, argument, science and explanation has been a massive bust).

The conceit of Milton Friedman and Paul Samuelson et al was that statistics and formally consistent math "models" would compel assent and agreement -- you could force people to the same policy conclusion by putting everything into math and numbers.

Friedman actually wrote a letter to Hayek insisting math and numbers could settle policy differences, and this proved the superiority of his economic science over Hayek's (a scientific strategy which Friedman patently didn't understand).

"Explaining things is hard. Persuading folks with a different understanding of things is even harder."

One of the things I sometimes do, but not often enough, is to first try to see where the other person is coming from. Then explain his own thinking in your own words. Once you have shown you have seen things through his eyes, you can then explain your own perspective. Not easy though.

Nick,

In their defence, prices rising due entirely to shifts in metaphysical supply and demand curves yet without any increase in the volume of physical transactions does sound rather mysterious.

Also, I guess MMT is hostile to the idea of inflation expectations in general.

vimothy: it is indeed. Plus, as one of the commenters said, in effect: "So, but somebody must be actually setting the price. And how would that somebody even *know* that demand had increased, unless he actually saw an increase in actual sales?"

Which is correct. There is no mythical Walrasian auctioneer observing supply and demand and adjusting price to equalise supply and demand in meta-time before trade is allowed to actually take place. (Some organised financial markets and auction markets aside).

Firms typically set prices for their products. And a firm won't know that demand has increased until *after* it sees a sustained rise in actual sales (especially if customers drop in at random).

Disequilibrium macro was following this line of reasoning, in the early 1970's. But, as Arnold Kling says, it was Theory X. It got swept away by the flood of theory Y, equilibrium theory, which just ignored this problem. And New Keynesian macro cut this Gordian know by just assuming firms set prices every k periods.

You cheated, and copied the wrong answer! It is "Sao Paulo", not "Sao Paolo".

Oh No! And I failed Geography too! Apologies to all Sao Pauloians (???)

Original post now edited for spelling.

Nick, your model with inflation expectations allows for infinite prices without a single transaction occurring. How do you realize a price without any transactions?

Sergei: not quite. I assumed the volume of transactions is the same before and after. Not zero.

But, many representative agent models do have prices, and zero transactions!

Technically we can define the "price" as the price that equalises supply and demand, even if quantity supplied, demanded, and transacted are all zero at that price. So it is never observed. It's a model, not the world.

Yes, and it isn't really in the spirit of the challenge, which is non-theoretical.

I was thinking about an analogy to a self-fulfilling currency crisis style model (since the mechanism so intuitively obvious), but I suppose it suffers from similar drawbacks...

Nick, no, you assumed that from one equilibrium to another there is a change in expectations, e.g. prices double. But your model allows for not just one doubling of prices but also doubling of doubling and so on until we reach our new equilibrium at infinite prices. Regardless of actual transactions, infinite prices are not possible because prices are finite by definition.

In other words your model does not depend on prices at all.

Sergei: that's why I cheated!

The original question was "what happens if there is an exogenous increase in the expected rate of inflation?"

I changed it to "What happens if there is an exogenous increase in the expectation of next year's price level?"

Starting in equilibrium, if the expected price level for 2011 increases, from one exogenously fixed level to another, that will cause the equilibrium price level for 2010 to rise by a finite amount.

If I had tried to answer the original question, I could not have assumed the nominal interest rate stayed constant. Because if I had, the current equilibrium price level would explode to infinity, as you say. I would have needed to invoke some sort of Taylor Rule in the monetary policy reaction function, or else hold the money supply fixed, or have some other nominal anchor.

So I cheated, to keep it simple(r).

I'm reading the comments again, and I'm reminded of a thought I had in the past.

Specifically concerning:

----------
Rogue Economist posted at Winterspeak:

If you are worried about looming inflation, you're less likely to eat out. You need to preserve your diminished purchasing power for the basic necessities. You will trade down to less costly substitutes. You will repair stuff rather than buy new ones.
----------

Winterspeak replied favorably to that, so I think that's the kind of answer he expected apart from 'do nothing'. It sounds to me like they're misusing microeconomic theory because they're thinking that an increase in prices, the inflation, automatically means a decrease in demand. That's the law of demand.

I argued with someone in the past and he seems to have thought the same. To prove that deflation was a good thing, he repeatedly cited the decrease in computer hardware prices in Silicon Valley. He then took that and concluded that deflation stimulates demand by lowering prices. By looking at it like an elementary microeconomic problem he completely missed most of what we know about inflation and deflation. Inflation is something that comes from multiple agents conducting successive transactions with multiple goods: even if our Brazilian did not generate inflation, history has shown that someone else obviously did. It sounds to me like an easy mistake to make if you're not an economist.

[very lightly edited for civility. NR. Keep it civil please.]

"And we make it easy for ourselves by using simple models with a very limited class of goods and assets. What we do looks a bit like cheating, for anyone who cannot or will not abstract from the world in all its complexity."

That's a bit unfair. The heterodox folk have models, too. Everyone abstracts away from the real world, it is how we think about the world.

You can make a good case that the flow-of-fund models are more accurate than DSGE models. They predicted this crisis, and do a better job at predicting macro time series such as employment and inflation. They are more "useful", but are rejected because they don't have utility maximization.

But what are the micro roots of NK models?

They aggregate preferences, which you can't do unless there is a dictator. So they are dictator models, with the same outcomes as social planner outcomes.

You don't have multiple simultaneous micro transactions giving rise to new macro effects, you have a single actor choosing to consume or save, and the outcome of this choice is supposed to determine economy-wide allocation of capital investment and consumption in each period.

It would be as if you tried to explain ocean waves by looking at a single molecule of H20, ignoring exchanges of energy and momentum in which one molecule in one state interacts with a different molecule in a different state. In that case, even mystical models that talk of angels "pushing" on water would be more accurate representations of the aggregate phenomena.

In statistical mechanics, you do not obtain the result that the aggregate behavior can be described by the same hamiltonian as that of the individual element.

At the same time, there is no allowance for asset bubbles. When you save, you increase the (real) capital stock by the exact amount saved. It's magic. Therefore there are no (financial) asset bubbles, there can only be an excess of real capital or an excessively low real interest rate. But between 1997 and 2007, real U.S. household net worth (which is the value of all the capital stock) increased by 170%, even though 1/(real FedFunds) increase by 120% and net investment, as a share of GDP, was less in the period 1997-2007 than the post war average. That isn't possible if households, in aggregate, save by increasing their holdings of real capital by the amount saved.

Because it isn't possible, there are no asset bubbles.

So you have to decide which model you trust more in describing the current predicament. The structural model without utility maximization but with a financial sector, or the dictator model without asset bubbles or a financial sector?

Reasonable people can disagree. Just as some do not believe in any of the predictions of the structural model, others can choose not to believe in any of the predictions (and specifically, the inflation predictions) of the barter-based dictator model. But this shouldn't be framed in terms of wise experts needing to patiently explain non-intuitive results to non-experts who don't bother with models.

Nick Rowe said "Macro is hard. And we make it easy for ourselves by using simple models with a very limited class of goods and assets. What we do looks a bit like cheating, for anyone who cannot or will not abstract from the world in all its complexity."

To me this is the fatal flaw of economics. The discipline has delusions of adequacy. Newton's model of the solar system wasn't perfect. If we proposed sending a rocket to Jupiter using Newton's model without incorporating Laplace's use of perturbation theory, because Newton's model is more tractable, we would rightly be laughed at.

Complexity is a defining characteristic of our economy and society. Assuming it doesn't matter is cheating, but that's not the bad part. The bad part is that the model error is not always explicit or obvious. Take the Euro as an example. I'm a not old enough to remember the Euro currency debates, but my understanding is that Mundell Fleming was used as a justification for creating the Euro. I don't think it would be terribly hard to use the very same model today to argue in favour of breaking up Euro.

I don't ascribe nefarious motives to any economist who used or uses Mundell Fleming to discuss the Euro, but perhaps we should take a step back and reconsider having high confidence in any recommendation from a discipline that gave out a Nobel Prize for a model that argues with itself?

Blogging is easy if you were educated at Chicago and always assume perfect competition. (Of course one's ideas may not seem plausible.)

Interesting to read your last sentence Joel, because it is a nice expression of what I wanted to say too.

"...a model that argues with itself". In a sense, all equilibrium models argue with themselves: because there are always competing effects in opposite directions - that is, of course, why we get an equilibrium.

And even apart from the forces which generate the equilibrium itself, the nature of the trade-offs in economics around an equilibrium point means that in any interesting question, there are very likely to be competing effects to disentangle.

A well-known example is the old debate about tax rates and supply of labour: will a higher rate of tax (therefore a lower net wage received by workers) reduce the supply of labour? Purely from the viewpoint of price, it will. But from the viewpoint of income, it might not. Perhaps workers will increase their supply of labour to maintain income at a target level. Perhaps they'll reduce it because they're not being paid as much for each hour. Economic theory doesn't tell us which effect is stronger, it simply helps us identify both effects (and in this case, helps us see whether the effects derive from the average or marginal price).

In Winterspeak's story, we have a similar confusion. It seems quite intuitive to suggest, as Winterspeak's commenter does, that "If you are worried about looming inflation, you're less likely to eat out. You need to preserve your diminished purchasing power for the basic necessities. You will trade down to less costly substitutes. You will repair stuff rather than buy new ones."

If we consider the person in the problem to have a fixed quantity of cash and no other income, with a requirement to make ongoing expenditure on perishable goods, then this is quite a logical response. Or, imagine a pensioner with a non-indexed pension.

Implicit in Nick's argument (and the mainstream macro argument) are two intertemporal simplifications, one of which must be satisfied to make the model work: that the representative agent possesses not just cash but income, which is expected to rise with inflation; and that most goods can be purchased this year and used next year.

If these assumptions are not satisfied, or not satisfied enough, then it is quite possible that the effect of inflation on wealth (or lifetime income) will dominate its effect on intertemporal substitution.

Again, economic theory can tell us both possibilities, but not which one is stronger. Empirical data (or sometimes common sense) is needed to answer that question. And I think Nick beats Winterspeak on common sense.

At least outside of the Walrasian barter auction room...

"Implicit in Nick's argument (and the mainstream macro argument) are two intertemporal simplifications, one of which must be satisfied to make the model work: that the representative agent possesses not just cash but income, which is expected to rise with inflation; and that most goods can be purchased this year and used next year."

But this is exactly the issue in a de-leveraging scenario.

Even if the economy contains only identical consumers, it is not true that the sum of each household's income statement is equal to the income statement of the group. The income statement of the group will be national income, but national income will in general be less than the average income of the representative household, even if all households have identical incomes and identical preferences.

The difference between the sum of individual's incomes and national income will be reflected by a growth in debt, which augments the incomes of individual households but disappears when you aggregate the individuals to get the national income.

When the economy is de-leveraging, then for a given level of national income, household incomes are declining due to the decline in borrowing. And this puts downward pressure on prices simply as households are constrained by their decreasing incomes.

And in that environment, it's unlikely that an increase in expected future prices will be credible in the first place, or if it is credible, that the individual household will have enough income to actually cause more inflation. The latter situation will only occur if the expectation of rising prices causes the de-leveraging to reverse.

The income/wealth effect trade off that you normally see is augmented by an overall income decline during a de-leveraging period, and this can seem -- if you are ignoring the effects of debt -- as if wealth effects are suddenly dominating the income effects. They are not, but each household's income is declining faster than national income, and this is a source of downward pressure on prices in addition to the normal trade-offs.

RSJ: huh!?! Borrowing does not contribute to income at the household level. Income is usually discussed in real-terms not nominal. So while net borrowing (against the CB) changes nominal income, it always deflates away in an evenly rotating economy once we consider the process in real-terms.

For similar reasons, net borrowing cannot be declining unless the CB is contracting the money supply, in which case, yes inflation is likely to be declining or negative.

Now what does happen is that the real interest-rate can become distorted. Artificially lows rates leading to a surfeit of investment and accelerating real production and artificially high real rates leading to a deficit of investment and decelerating real production.

"huh!?! Borrowing does not contribute to income at the household level."

Borrowing is a key driver of income.

Here is an example.

Suppose your economy has two people, A and B, one firm, and one bank. Say the firm has an enterprise value of 2X, and is owned equally by A and B.

At the beginning of the period, A starts out with X assets. B starts out with a house that he purchased for $X some time ago, as well as X in firm assets.

A borrows 2X from the bank in order to buy B's house. B spends X on goods from the firm and saves the remainder of his income.

A doesn't buy any goods from the firm.

So the firm receives X in revenue and pays out X in the form of wages and capital income equally to both A and B (the firm does not do any additional investment in this period).

So what do our income statements look like?

A receives X/2 in wages and interest income and has no expenses.
He saves .5X in this period.

A's balance sheet, by the way, goes from X assets and 0 liabilities to having a house valued at 2X as an asset, .5X of deposits , X in firm stock, and a debt of 2X as a liability.

B receives X/2 in wages and interest income, as well X from the disposal of the asset, for income of 1.5X. He consumes X and saves .5X in this period.

B's balance sheet goes from having a house valued at X, firm stock valued at X, and no liabilities, to having 1.5X in deposits together with firm stock valued at X. B's balance sheet increase by the amount that he saved -- .5X.

The bank's balance sheet goes from 0 to 2X in assets (the loan) as well as 2X in liabilities (the deposits).

But what happens at the level of national income -- where you are picking your "representative" agent to allocation a portion of national income towards investment and a portion towards consumption based on his preferences?

At the national level, there is no investment. The house already existed and was not improved -- no residential investment occurred. And the firm did not invest, either.

Therefore national income is just B's consumption, or X.

But the sum of individual's incomes is 2X, as A received .5X in income and B received 1.5X in income.

Therefore the capital gain can be realized as income at the level of individual households, but does not appear in NIPA at all.

It is invisible to the NK model, as the NK model does not allow for an increase in asset values without this being matched by an equal increase in actual investment. NK models -- and all "non-heterodox" models, rule out asset bubbles as a result of how they model the law of motion for capital.

Therefore they can't predict the bursting the asset bubble, or what remedies should be taken once the bubble bursts. They don't understand that as soon as A stops borrowing, then this puts downward pressure on prices.

"So while net borrowing (against the CB) changes nominal income, it always deflates away in an evenly rotating economy once we consider the process in real-terms."

Pure fantasy.

"For similar reasons, net borrowing cannot be declining unless the CB is contracting the money supply, in which case, yes inflation is likely to be declining or negative."

The flow of funds says otherwise.

"Now what does happen is that the real interest-rate can become distorted. Artificially lows rates leading to a surfeit of investment and accelerating real production "

In this example, there was no excess of real investment, as real investment was zero. According the BEA, actual investment declined over the course of the last cycle, yet we had a pretty big asset bubble, and outsized incomes to boot.


"Your answer must use words only, with no diagrams or equations."

That is not constraint in blogging unless self-imposed.

Adam P@10:19AM:"it was clear in an exchange between Winterspeak and Rebel Economist"

Not Rebel but Rogue, I suppose.

sorry for offtopic but is there an RSS feed for all comments on this blog?

Sergei: go to the top of the list of comments here, immediately underneath where it says "Comments". It says: "You can follow this conversation by subscribing to the comment feed for this post."

RSJ: "That's a bit unfair. The heterodox folk have models, too. Everyone abstracts away from the real world, it is how we think about the world."

That's not really what I meant. Yes, the heterodox folk have models too. And yes, we all abstract, all the time.

What I meant is this. When we build a model, especially a macro model, we simplify all the bits. Someone who is very familiar with one of those bits, say someone working in sales, will look at that bit and say it's totally wrong. "We can't just sell what we want to; it takes people like me to make a sale, and your model leaves that out!"

"But what are the micro roots of NK models?

They aggregate preferences, which you can't do unless there is a dictator. So they are dictator models, with the same outcomes as social planner outcomes."

Yes, NK models (typically) are representative agent models. But no, they very definitely do not have the same equilibrium as the social planner would choose. Some RBC models do (all the early ones did).

And you can only aggregate preferences if preferences are identical homothetic, not dictator.

RSJ: You are not understanding NK models correctly.

1. Even if all agents are identical, you can get a difference between micro effects and macro effects, and NK models have that. The fallacy of composition does not disappear if all agents are identical.

2. Simple NK models do not allow bubbles in asset prices. This has nothing to do with whether they allow borrowing and lending (which they do allow). It's because simple NK models have the capital good the same as the consumption good. If the technology is, for example: deltaK+C=Y=F(K,L) then this says the marginal rate of transformation of the capital good into the consumption good is equal to one everywhere. And this pins down the price of K to equal the price of C.

There's a production process, Y=F(K,L). So K and L produce Y. At the end of the assembly line, where Y comes out, there's a switch. Flick it one way, and the Y comes out as I. Flick it the other way, and Y comes out as C. What's more, in very simple models it's reversible. You can eat the machines. That's why the price of machines must always equal the price of cake.

Nick, I rather meant _all_ comments from _all_ posts in _one_ RSS feed. I find it very cumbersome to subscribe to every post every time. But I also find comments here quite interesting to miss an opportunity

Sorry RSJ: Perhaps you do understand my point 2 above. That's maybe what you meant by "NK models -- and all "non-heterodox" models, rule out asset bubbles as a result of how they model the law of motion for capital."

But that's not an issue of "heterodox" vs "orthodox". It's whether you have a simple production function or have a more complicated one with 2 goods. Wlarasian GE theory is very orthodox, but has n goods.

"It is invisible to the NK model, as the NK model does not allow for an increase in asset values without this being matched by an equal increase in actual investment"

if that's the case, thank goodness

enough of a mess between accounting and economics

keep the NK model NIPA consistent and adjust for asset values from there

the important point is that capital gains in aggregate are not available for the purchase of GDP

"the NK model does not allow for an increase in asset values without this being matched by an equal increase in actual investment"

this statement is patently false.

Adam: Isn't it true in a simple NK model (actually, in any simple model, NK or not), that has the technology Kdot + C = Y(K,L) ?

But sure, there is no problem in building an NK model with I and C being two different goods. Or even adding land into an NK model. It's not an "essential" feature of NK models.

anon: "the important point is that capital gains in aggregate are not available for the purchase of GDP"

Agreed. But asset prices may affect the demand for GDP, in much the same way that interest rates affect the demand for GDP. (Not that you would deny this, presumably, but just to remind ourselves.)

"Isn't it true in a simple NK model (actually, in any simple model, NK or not), that has the technology Kdot + C = Y(K,L) ?"

No, anytime the natural interest rate changes the value of all assets, bonds and equities, changes in exactly the right way to keep investment unchanged.

If the natural rate of interest has fallen then all asset prices increase with no change in investment.


"the important point is that capital gains in aggregate are not available for the purchase of GDP"

You are confusing an ex-post accounting identity with an ex-ante constraint, as well as confusing the balance sheet of the individual with the balance sheet of the group.

In the example I gave, when you aggregate A and B's balance sheets, then there is no (net) capital gain. The house increases in value, but net indebtedness increases by the same amount.

In terms of NIPA, savings = 0 in this economy. But each actor saves .5X.

Ex-ante at the level of the individual consumer, investment income is an important contributor to the individual's income, and income from any source is available for any use, including the purchase of goods which contribute to national output.

At the end of your accounting period, once you've aggregated all the transactions for the group, and have aggregated the balance sheets of the group, then you then classify all the expenditures that occurred on final output as consisting of national income over that period, and everything else you classify as a balance sheet adjustment.

That is an ex-post classification, not an ex-ante constraint.

If B would have elected to spend more of his gain on consumption, then national income would have increased. National savings would still be zero, but the sum of individual's savings would decrease.

And this conflation between national income, national savings, and individual income and individual savings is at the heart of the matter.

What drives economic activity in a model with micro roots?

Do you first aggregate the income and then apply an euler equation to national income, or do you apply the euler equation against individual's household income and savings, and then aggregate to get the law of motion?

It makes a material difference, even if all consumers are identical. This has nothing to do with making abstractions or simplifications, it's just a question of improper aggregation.

“You are confusing an ex-post accounting identity with an ex-ante constraint, as well as confusing the balance sheet of the individual with the balance sheet of the group”

2nd point first – I said “in aggregate”, so I’m not confusing anything

1st point – it’s both an ex post identity and an ex ante constraint.

It’s an ex ante constraint because for any prospective accounting period, expenditure will be equal to income, in aggregate. Any individual who consumes from capital gains will force some other individual to save from income; that forces the expenditure/income identity ex ante, which means capital gains are not a part of it in aggregate.

Capital gains are not a part of macroeconomic income.

P.S.

Capital gains are not paid from producers to the factors of production.

That's why capital gains are not part of macroeconomic income.

"I said “in aggregate”, so I’m not confusing anything"

In aggregate, capital gains are ZERO in the example. In aggregate, "the economy" is NOT obtaining national income from capital gains. So you must be confusing individual with aggregated balance sheets, otherwise you wouldn't be complaining that the individual capital gain cannot add to national income.

"Capital gains are not paid from producers to the factors of production."

Never said that.

"Capital gains are not a part of macroeconomic income."

Correction -- and really, for the last time:

"Capital gains OF THE AGGREGATED BALANCE SHEET are not part of NATIONAL income".

That is a true statement, which does not contradict the following true statement:

"Capital gains OF THE INDIVIDUAL *can* contribute to NATIONAL income"

"It’s an ex ante constraint because for any prospective accounting period, expenditure will be equal to income, in aggregate. Any individual who consumes from capital gains will force some other individual to save from income;"

Then how do you explain the example? You are assuming -- and this is where you are confusing ex-post with ex-ante -- that *individuals* are paid with some share of *national* income.

But it's really the other way around.

National income is *unknown* until all the micro decisions are made.

It cannot be an ex-ante constraint on individual's incomes.

In the example, once B sold the house, he had complete freedom to spend all of the gain on consumption, some on consumption, or none on consumption. Based on B's choice, A's savings would adjust, and national income would adjust.

And this consumption decision completely determines national income ex-post.

B obtained the gain *not* from A's saving, but from A taking out a loan to buy the house.

The bank created the money that went to B.

The corresponding deposits were then distributed among both A and B based on their own consumption/savings choices.

The development of your argument is not consistent

You interjected at my:

"the important point is that capital gains in aggregate are not available for the purchase of GDP"

that statement is true and you've done nothing to disprove it

you're confusing aggregate demand with the identities that constrain the equivalence of output and income; the identities hold ex post and ex ante even if capital gains cause an increase in aggregate demand at the micro level

you have a minor case here of Krugman's "Dark Age of Macro" infliction

Look, this is really simple.

If you calculate each individual's income, and add that up for all the individuals, you will not (in general) get national income. The reason is that due to credit growth, individual's incomes may increase more than national income. And if credit is contracting, then they will decrease faster than national income.

I gave a very simple example illustrating this, and you haven't pointed out any errors with it. No one has.

Obviously the expenditures of each individual, on all sources, are equal to individual income.

But the sum of individual expenditures is not national output, and the sum of individual income is not national income.

If you agree to the above statements, then we have no argument.

In that case, it's another example of pointless disputation.

I'm beginning to suspect that this is the case. I should have known, of course.

But if I'm wrong, and you really believe that the sum of individual incomes must equal the national income, then I would like to see your "proof", as well as where the example I gave makes an error.

If you define "income" as "income from the sale of newly-produced goods", and "expenditure" as "expenditure on newly-produced goods and services", and include accumulated unsold inventories as pert of "expenditure" (you sold it to yourself), and add in net exports of newly-produced goods and services,...... then all three are equal. (Unless I've forgotten something). Ex post, but not ex-ante.

Nick @ 6:35,


"If you define "income" as "income from the sale of newly-produced goods", and "expenditure" as "expenditure on newly-produced goods and services", and include accumulated unsold inventories as pert of "expenditure" (you sold it to yourself), and add in net exports of newly-produced goods and services,...... then all three are equal. (Unless I've forgotten something). "


The problem with this is that you cannot make up your own definition of income.

An income statement is the Profit and Loss over a period, showing revenues and expenditures according to some basic conventions that involve properly recognizing revenues in a consistent and rational way.

That is the definition.

You can't change the definition.

Saying that income is the income from sale of newly produced goods is the result of applying this definition to the national level, where one firm has revenues of X_1 and expenses of X_2, which accrue as revenues to another firm, which has X_2 as revenues and expense of X_3, etc.

So the total national income is a telescoping series in which you can deduce that P&L for the nation as a whole consists of, on the revenue side, "income from the sale of newly-produced goods. "

But that's not the definition of income, it's the algorithm for aggregating incomes of disparate groups within an economy.

You cannot apply that algorithm to an individual actor -- it doesn't make sense. You would get all sorts of violations in which expenditures are not equal to revenues for each individual actor. For example, in the example I gave, B would have expenditures of X, and income of .5X, which necessarily means that B is dissaving and that should show up on his balance sheet. But B isn't dissaving. It's because you are using an inconsistent definition.

Both in the case of a group of individuals and in the case of a single individual, you apply the same definition of income.

This is just common sense.

In any case, you want a model that has micro-foundations, right?

If a household is receiving wages and capital income, then their income statement has a specific meaning that is not equal to "income from the sale of newly produced goods". The household may not be producing much in the way of newly produced goods. Nevertheless it will not be selecting it's consumption and savings plans based on that (inconsistent) definition of income, but based on it's actual expected income.

In other words, whenever you want to calculate an income statement or balance sheet, you have to define the accounting boundary.

If I take $1 out of my left pocket and put it into my right pocket, that is not income.

If you take $1 out of your left pocket and give it to me, then it is income for me and an expense for you.

But then if we were to consolidate our balance sheets, it would stop being income. My income would cancel with your expenditures.

You would need to subtract out the revenues A receives from B if both A and B are being consolidated.

That is how you "prove" the telescoping result that national income is the revenues from newly produced goods. It's not a definition of income, but an algorithm derived from the actual definition of income, which is just revenues.

But when looking at euler equations, if I expect to get a dollar from your pocket, I will absolutely take that into account when forming my consumption and investment plans.

So the appropriate scope for balance sheet consolidation in a macro model is the economic decision maker.

Even if you select a representative agent, it is not the case that the accounting boundary is for the entire household sector, or even for the entire private sector.

The appropriate boundary remains that of the individual household, and you just happen to be using a contrivance in which each household is identical. That is not the same as assuming some sentient "household sector" that is optimizing.

By the way, the above observation is a problem I have with the MMT view of the importance of "net financial assets" for the private sector as a whole. No one makes decisions based on net financial assets of the private sector. But that's a whole separate story :)

"A borrows 2X from the bank in order to buy B's house. B spends X on goods from the firm and saves the remainder of his income."

For starters, that 2X is not income. It's the value involved in the exchange of capital assets.

"But the sum of individual expenditures is not national output, and the sum of individual income is not national income."

Do you have any idea how absurd that is?

Nick -

"If you define "income" as "income from the sale of newly-produced goods", and "expenditure" as "expenditure on newly-produced goods and services", and include accumulated unsold inventories as pert of "expenditure" (you sold it to yourself), and add in net exports of newly-produced goods and services,...... then all three are equal. (Unless I've forgotten something). Ex post, but not ex-ante."

Right. But it holds ex ante as well. It's an accounting identity that must hold for any future accounting period. That doesn't tell you anything about what the numbers are - just the relationship that must hold by identity between the numbers.

If you really think accounting identities won't hold for future accounting periods, I'd like to know why.

No, Anon,

"For starters, that 2X is not income. It's the value involved in the exchange of capital assets."

I didn't say 2X was income. You don't know the the B's income yet, because it depends on how much wages and interest income each person receives, which in turn depends on how much B and A decide to spend (in this simple model). So B's income will depend on the choices made by A.

But B bought the house for X and sold it for 2X, so he does have X of investment income. He will get more income from wages and interest, of course. How much more, we don't know yet. But he chooses to spend X on consumption and no more. This is perfectly plausible (and there is no error here).

I'm sorry if you find this too confusing or "absurd".

anon: "If you really think accounting identities won't hold for future accounting periods, I'd like to know why."

Because of monetary exchange. Desired aggregate expenditure on newly-produced goods and services is not *identically* equal to expected income from the sale of newly-produced goods and services. They are only equal in equilibrium. Say's law is false.

continued; "ex ante" is a crude way of saying "planned" or "desired", or, in normal microeconomic language, "demanded". Quantity demanded is not identically equal to quantities supplied, or to quantity transacted. They are only equal in equilibrium.

"But B bought the house for X and sold it for 2X, so he does have X of investment income"

That is completely wrong.

Sorry, but there's no point in continuing this. You simply don't know what the definition of income is, including investment income.

Nick,

Think about it. Accounting identities hold for all periods - past, present, and future.

It's got nothing to do with Say's Law or planned or desired this or that. Those things end up determining the numbers that make up the accounting identities, but not that the identities are a constraint for any given measurement period - past, present, or future.

This is a variation on what Krugman was talking about in "Dark Age". That was all about Chicago guys misinterpreting accounting identities, although in a different way than here.

RSJ,

Re the definition of income - I find it absolutely staggering that you would consider the sale of a house as "investment income". I sincerely hope somebody else comes along and corrects you on this, since you write a lot around this sort of thing on these blogs and clearly have a lot to offer.

Accounting identities always hold. Income statements are a way of aggregating transactions that occur over a period of time, say Q1 2011.

If B had decided to spend X/2 on consumption in January, and no other transactions occurred in Feb. or March, then national income would have been X/2 in that quarter. You will need to wait until April 1 to find out. That is what I mean by ex-post.

In general, under the assumptions of the model (A is deciding go into debt to buy B's house for 2X, triggering investment income of X, of which B spends a portion, C; the bank is willing to finance A's purchase; A decides to save all of his income during the quarter; firms do not invest) then for any number C between 0 and X, we have A's income = A's savings = C/2. B's income = X + C/2, of which C is spent on consumption and X - C/2 is saved, and national income = C.

Therefore A's income + B's income = national income + X, where X = B's investment income = the net amount of new money created by the bank during the course of the accounting period = the amount of financial savings realized by the sum of A, B.

The point here is not to argue that economies always follow these rigid rules, but to highlight that applying the euler equations to the individual will yield different results than applying the same equations to the aggregated group.

If you just look at the national income, you only see C -- X disappears. The role of debt growth in fulfilling financial savings demands disappears, as does the pain of de-leveraging.

Blogging is hard, Nick, but I admire your courage and creativity in constantly coming out with your thoughts to a public audience. This is how ideas get refined, so thank you for your efforts in this debate. The important thing is not to have the right answer all the time, but to be involved in the most fruitful efforts in the search for answers.

My take on this is close to RSJ’s, I think. Debt does make a difference. Inflationary measures can be hindered in a recession economy with a lot of debt, because the indebted that are undergoing deflation, i.e. balance sheet recession, will not be induced to make more purchases if paying off the debt is already eating up much of their income. They will hope that the non-indebted ones will come to the rescue, and be induced to consume more via increased inflation expectations. But then, the non-indebted have to be convinced that the ongoing balance sheet recession among the indebted will not counteract any of the inflationary measures.

Anon,

Bombast is not helpful here.

The investment income is the income from the disposition of the asset, which was purchased for X and sold for 2X, for investment income of X. Not 2X, as you keep trying to suggest. I never said the investment income was 2X.

If you want to learn more about investment income, please consult any accounting text.

It is not staggering at all.

It is not even confusing.

Anyone with a rudimentary knowledge of accounting will be able to help you understand this concept, if you remain confused about the possible existence of investment income.

It doesn't matter if it’s X, 2X, or zero. The sale of a house for X, 2X, or zero does not constitute income. If there is a capital gain, it is a capital gain, not income.

Income is rental on a capital property (such as a house) – not the sale of a house or the capital gain on a house.

If tax authorities include capital gains as income in whole or in part, that’s for purposes of calculating taxes, not to define economic for economic purposes. The fact is that capital gains must be excluded from the calculation of economic income at the micro and macro levels as well as NIPA at the macro level.

You can call it income if you like, but it just confuses things and doesn’t contribute to the explanation of economics – just muddies it. At the end of the accounting period, capital gains are excluded from national income and nothing of any economic consequence is gained by defining it as income at one level and not income (properly) at the aggregated level.

Demonstrating the economic effect of capital gains (e.g. increase aggregate demand) from selling a house doesn’t require that you classify it as income and conclude that there is some asymmetry between micro summation and macro.

Any expenditure of capital gains at the micro level forces an equivalent amount of marginal negative saving somewhere else. It must be so, because macroeconomic income cannot exceed the totality of income that is generated excluding capital gains. If you think otherwise, you'd have to include the capital gain on the house as part of GDP, in order to balance your accounts at the macro level, and including capital gain on a house as part of GDP is also nonsense.

anon: "Nick, Think about it. Accounting identities hold for all periods - past, present, and future. It's got nothing to do with Say's Law or planned or desired this or that."

I know that accounting identities always hold, past, present, and future.

But the *conventional* meaning of "ex ante" in economics *is* "planned" or "desired", or "demanded". And that has to do with Say's Law.

RSJ and anon: there is more than one way to do accounting. And there is more than one definition of income. The National Income Accounting definition of income (which is a convention among most economists) is very different from the way my tax accountant measures my income. For example, NIA excludes capital gains, and my accountant includes *realised* capital gains.

There is no one right definition of income. Different definitions may be useful for different purposes.

Thanks rogue!

"Any expenditure of capital gains at the micro level forces an equivalent amount of marginal negative saving somewhere else."

I meant an equivalent amount of marginal saving somewhere else. Expenditure from capital gains at the macro level constitutes marginal negative saving (spending from zero income).

Nick,

What's your interpretation of Krugman's Dark Age post?

(It certainly has something to do with the relationship between accounting identities and economics):

http://krugman.blogs.nytimes.com/2009/01/27/a-dark-age-of-macroeconomics-wonkish/

anon: Yes, it does. Here's PK's Dark Age post in a nutshell:

"Quantity bought is identically equal to quantity sold. That's an accounting identity. Some people get confused into thinking that that means that quantity demanded is identically equal to quantity supplied".

BTW: your only two mistakes in the above exchange, as far as I can see:

1. You misused the words "ex ante". (You are not alone, and those are very misleading words.)

2. You forgot that non-economists may use different accounting conventions.

In my various arguments with those coming from an accounting background, the mistakes they most frequently make are:

1. They think that accounting identities are statements about the world, as opposed to a set of consistent definitions of words.

2. They have difficulty distinguishing "quantity demanded" from "quantity bought", and "quantity supplied" from "quantity sold".

I'm cross posting between two posts now.

[No worries. I have unpublished the other one. NR]

Here's a repeat:

BTW, here’s how I’d construct a much simpler example to illustrate these points.

Suppose we have a two person economy with macro output (GDP) and income of X.

Each person’s income is X/2.

Suppose I sell my house to the other person for X.

Suppose I purchase X – the entire economy’s output - and it is all consumption goods, no investment goods.

Then my consumption is X and my dissaving is (X/2).

The other person’s consumption is zero and his saving is X/2.

Macro consumption is X and macro saving is zero.

The proceeds from selling my house arguably have boosted my contribution to aggregate demand, with the result that consumption has been skewed in my direction. This is entirely separate from the issue of how to account for the sale of my house. I have accounted for it as a capital transaction – which is correct in economic terms. Whether there is a capital gain in accounting terms is irrelevant, although in economic terms it may have affected my psychology and my behaviour in terms of my contribution to aggregate demand. It is also irrelevant whether the buyer has borrowed from the bank to purchase my house or has drawn down prior asset saving.

With reference to the accounting, it makes no sense to me that anybody who favours a coherent discussion at the macroeconomic level would deliberately force through inconsistent definitions at the micro level. That’s how I approach the subject of accounting and choices about accounting definitions.

I think the confusion here stemmed from whether to make clear distinction between flow and stock. Ordinary people do; Economics doesn't.

If you contrast conventional concept and economic concept in regards to asset, it goes like this:

Conventional concept : Asset(Stock)
Translated into Economics : Initial endowment plus savings

Conventional concept : Asset(Stock) Price
Translated into Economics : Discounted Present Value of future cash flow

So, when the asset price changes, economics treats it as change in future cash flow and/or change in discount factor. Thus, it can be incorporated into intertemporal model such as Euler equation. If you increase current expenditure due to capital gain, in economics, it just means that you spend the future cash flow in advance.
On the other hand, ordinary people perceive the asset price change as change in the evaluation of the asset. So, to them, capital gain is something like gift from heaven, and is not so different from other forms of income.

NIck,

"The National Income Accounting definition of income (which is a convention among most economists) is very different from the way my tax accountant measures my income. For example, NIA excludes capital gains, and my accountant includes *realised* capital gains."

Certainly there are different accounting conventions. But in my example, the gain was realized. I was not assuming that merely the increase in market value of the asset would generate income for the asset holder or would contribute to aggregate demand.

Moreover, the NIA accounting is really not fundamentally different from the accounting that would be used by an individual firm. Income = Revenue, but it is revenue received as a result of transactions that cross the accounting boundary of the firm. The transfer of funds from one subsidiary to another does not count as revenue for the firm.

That's all it is.

Imagine, if one firm buys another, then what would be the revenue of the combined entity? Suppose you only had access to the income statements of both. You would need to say that the income of {A,B} = Income of A + Income of B - the expenditures of A on B - the expenditures of B on A. You subtract in order to avoid double counting. You can argue that the result is "final output" _for_ the combined entity.

Now if you keep consolidating income statements to include every production unit, the net result is national income. It's still just "revenue" and is defined as revenue, but if you want to calculate the revenue, given only the income statements of the constituent production units, you would need to subtract out all the income received by the unit that was an expense for some other production unit.

Now obviously you cannot argue that the individual firm (or household) is going to optimize over this (reduced) income.
But that's just what the models do!

The reason why the national income accounts don't need to count income received from asset sales or government transfers. It's not because gains from asset sales or transfers are not income. They are income for the individual, but because these are transactions that do not cross the accounting boundary when the boundary constitutes every production unit.

But just because your model is using a representative agent does not mean that the accounting boundary expands to include every production unit.

The accounting boundary needs to be the same as the decision maker.

And continuing along these lines,

The key point of dispute here, is that both you and Anon are making is that the increase in income from the gain realized by B is equal to an increase in expenditures by A, so that in an economy containing both A and B, the sum of A and B's income will be unchanged. And therefore in an economy containing a single "representative" agent, you can safely ignore realized gains.

And this false!

The reason why it is false is because of balance sheet expansion.

When A borrows to buy B's house, that is not considered to be an expenditure on A's income statement because it does not subtract from A's net worth. Regardless of which accounting convention you are using, the purpose of income and expenditures is that when one is greater than the other, then net worth must change. That is why income is a non-cash measure of revenue.

But if you look at A's balance sheet, both liabilities and assets grow by the same amount. The acquisition of capital is not an expenditure for A, and so it is not the case that B's investment income from the disposition of the must be "paid for" by A's expenditure. The two items do not cancel.

"Suppose we have a two person economy with macro output (GDP) and income of X."

That's not how it works.

You don't know the aggregate income *until* the transactions have occurred. You are confusing (again) ex-post with ex-ante. Income is measure of revenue between two time periods. It is determined by individual behavior over the course of the time period. It does not determine individual behavior prior to the end of the reporting period.

Accounting constraints are nothing more than consistency constraints on models. They are not behavioral constraints, they are constraints on how you measure behavior.

"Whether there is a capital gain in accounting terms is irrelevant, although in economic terms it may have affected my psychology and my behaviour in terms of my contribution to aggregate demand."

"Macro consumption is X and macro saving is zero."

OK, making up your own terms is not helpful.

Households do not have "macro savings".

They have savings.

The "macro savings" is an attribute of the macro balance sheet, not the individual's balance sheet. There is no such thing as "macro savings" of a household.

The economically relevant factor is the income statement and the balance sheet of the individual, not that of the group.

The macro phenomena arise from and are determined by the behavior of individuals, not the other way around.

"Whether there is a capital gain in accounting terms is irrelevant"

No, it is not. If there was a capital gain, then your net worth has increased. An increase in net worth is not "irrelevant" in accounting terms. You can debate whether to use accrual or cash-basis -- my example uses a cash basis. If you are using accrual accounting, then the increase in net worth occured when the market price went up. But in any case, an increase in net worth is not "irrelevant" from an accounting sense.


"although in economic terms it may have affected my psychology and my behaviour in terms of my contribution to aggregate demand. "

No, an increase in net worth is not "psychological". It really means that you have saved. Your assets are greater than your liabilities. You have received income. Now you can debate as to which accounting standards you should use as to when you realize the gain. In my example, the gain was realized when the house was sold. But whenever you realize the gain, it is a real gain. It's not psychological. This gain does not come at the expense of a decline in net worth by someone else.

A's savings are not "paid for" by the purchaser of the house.

RSJ,

I’m amazed at your durability. Do you come with rust proofing and snow tires as well?

“Income = Revenue”

No its not. Not for a firm.

“you would need to subtract out all the income received by the unit that was an expense for some other production unit”

You’re confusing revenue and income for a firm.

“and so it is not the case that B's investment income from the disposition of the must be "paid for" by A's expenditure”

The sale of a property is not investment income for a household. Rent is investment income.

Anyway, you’re just making up your own accounting definitions here. You could make whatever point you’re trying to make by constructing a flow of funds statement without resort to definitions of income.

“You don't know the aggregate income *until* the transactions have occurred. You are confusing (again) ex-post with ex-ante.”

Why are you not allowed to simulate scenarios of future transactions in your world?

“It is determined by individual behavior over the course of the time period. It does not determine individual behavior prior to the end of the reporting period.”

It’s determined according to feasible outcomes that are as always constrained by accounting identities within the future scenario that is being simulated. If all output within the period by construction of the simulation has been purchased at time epsilon minus one second, somebody who hasn’t yet purchased may well be forced to save by accounting identity under consistent scenario construction. Production and consumption won’t be infinite just because the accounting period we specify is in the future.

“They are not behavioral constraints, they are constraints on how you measure behavior.”

They’re behavioural constraints whenever you simulate future behaviour for a future defined accounting period. You can’t simulate a feasible future outcome that doesn’t adhere to accounting identities. That constrains your assumption for future behaviour. The constraint is on the simulator, such as yourself.

“OK, making up your own terms is not helpful.”

LOL, pot!

Suppose I cave momentarily and allow you any definition of income you like, including any inconsistency of definition you choose between micro and macro.

Can you then summarize the point of economics you are making? And can you make that point independent of and beyond the accounting definition for income or anything else? That might be more productive.

Anon, you are the one arguing that households are receiving "macro income" and obtaining "macro savings".

That's inventing your own (inconsistent) accounting. Households receive income and save. They do not "macro save" or receive "macro income".

In terms of revenue versus income, this is going to depend on which standard you are using. It is the top line of the income statement according to most definitions; you are not really making a substantive point here by arguing that according to GAAP, you would make a distinction between operating income and total income. In NIPA accounting, you don't have this distinction, because in general they don't face the same issues that investors face in trying to determine the quality of income.

That's more pointless disputation, IMO, but I'll leave it at that.

"They do not "macro save" or receive "macro income"."

You're making up stuff I never said.

I constructed a simple example in which one person consumes all the output.

So you transpose that event to misleading generic vocabulary, because you didn't follow the specific example, and it suits your theme not to do so, I guess.

Anon:

"“Income = Revenue”

No its not. Not for a firm.
"

I must be missing something here. It's simple accounting isn't it ?

"Revenue" is the amount the firm realizes when it sells goods or services.
"Expense" is what is spent to produce goods and services.
"Profit" is "Revenue" minus "Expense".

On the firm income statement, "net income" is the amount of profit earned during a fiscal period. Is anyone disputing these well-known definitions ?

"
The sale of a property is not investment income for a household.
"
For the firm, a property sale transaction result would be recorded as investment income(or loss) in the non-operating part of the income statement (other revenues & income). Why should the household be any different ?

Is this macro vs. micro accounting difference ?

I grow 100 apples, eat 60, and sell the remaining 40 at $1 each, and buy $40 bananas at $1 each.

What's my income? $40, or $100? Do I sell those 60 apples to myself?

Who cares?

It really doesn't matter, as long as we define expenditure the same way. We get exactly the same answer when we maximise utility, a function of apples and bananas consumed, subject to either of the two budget constraints.

I earn $100, and my house appreciates by $20 over the year. I decide to stay in my house, and spend the $100 on food.

What's my income? $100, or $120? Do I sell my house to myself, to realise the capital gain?

Again. Who cares?

So to sum up, economics is not accounting?

It matters

- if you're aggregating
- if you're having a discussion around the economic effects of aggregation
- if you're using the vocabulary of income and logically related measures in that discussion
- if you want consistent logic and math for aggregation in that discussion
- if you want to have consistent communication in that discussion

In other words, given the interest of this blog in macroeconomics, and given that list of issues, it matters tremendously


anon: OK. Now we're getting somewhere.

So, your answer is that some definitions of income will aggregate up, and others won't. And it is useful/convenient/good insurance against errors/etc., if we choose a definition of "income" such that when we add up each individual's "income" we get aggregate "income", without switching definitions.

Makes sense.

Now, can you: give me examples where a definition of "income" won't aggregate up? That way, we can see what's at stake.

(I think both my definitions in the "apples" example aggregate up OK. What about in the "house" example?)

If we define "income" as "value added" (sales minus purchases from other firms) it aggregates up across firms. If we define it as "gross sales", it doesn't. That's one example. That's what you were arguing with RSJ about, right?

Nick,

If you include capital gains in household income, it won't aggregate up to aggregate income.

There is no payment from factors of production, just because a stock goes up in value.

So GDP won't equal aggregate income if you include capital gains.

anon: that's true, of course. But maybe we should re-define aggregate income to include aggregate capital gains? "Income"=GDP+Capital gains? Pros and cons of doing this?

P.S.

The logical reason to exclude capital gains from a uniform (micro and macro) definition of income is that capital gains are a function of the value of assets that have already been recorded in the past as part of GDP production. Income corresponds to the value that is being produced in the current accounting period. So there is a mismatching of purpose when you mix capital gains on old assets in with the value thats being produced by the creation of new assets.

Would you rather have GDP = income, or GDP = income - capital gains?

The comments to this entry are closed.

Search this site

  • Google

    WWW
    worthwhile.typepad.com
Blog powered by Typepad