There is no “paradox of thrift,” and even if there were it wouldn’t matter . . .
. . . for public choice problems addressing long run fiscal issues.
Matt Yglesias has a post on consumption taxes:
The great egalitarian political philosopher John Rawls wrote that he preferred the idea of a consumption tax to an income tax “since it imposes a levy according to how much a person takes out of the common store of goods and not according to how much he contributes.” In response to a very similar argument from Scott Sumner, a smart Steve Roth post replies that financial saving is not the same as saving real resources:
This makes absolutely no sense. If you forego a massage this week, or wait a few years to get your house painted, is the labor for that massage or paint job “saved”? How about this year’s sunlight “” the ultimate source of that labor power? Can you use it next week, or next year? Understand: services comprise 80% of U.S. GDP. And that’s before you even think about Apple and similar, with their just-in-time, on-demand supply chains “” when you buy it, and only when you buy it, they produce it.
If you don’t buy it, it doesn’t get produced.
I think that to understand the Sumner/Rawls view you have to remember that both are assuming that the economy is always operating at full employment. Rawls doesn’t specifically say anything about this, but it’s the only way to make his viewpoint make sense. Sumner writes extensively about business-cycle issues, however. One of his main themes is that a competent central bank can always guarantee full employment and that it always should guarantee full employment in part because the full employment macroeconomy of steady national GDP growth is one in which all these nifty neoclassical ideas actually work. So the way the story goes is that the people thrown out of work when you switch from consuming to saving will be reemployed in the production of capital goods. We all become thriftier, stop dining out so much and start cooking at home, and all those unemployment cashiers and waitresses get jobs building houses and manufacturing tractors. Thus society’s stock of capital goods does in fact increase through a big shift toward a higher savings rate.
I absolutely do not assume the economy is at full employment when advocating consumption taxes. And “financial saving” is a meaningless term, so I won’t comment on that. Saving is saving; it is defined in all the textbooks as the funds that go into investment. (You are free to have your own definition.) There are actually three errors embedded in the Roth/Yglesias critique:
1. There is no paradox of thrift, so saving doesn’t cause higher unemployment. Oddly, the most common error on this topic is exactly the opposite; most people think high saving/CA surplus economies steal jobs from low saving economies, a view that is equally wrong. Periods of higher than normal unemployment are caused by either NGDP shocks (bad monetary policy) or bad supply-side policies (think France/Italy/Spain.)
2. OK, most old-style Keynesians don’t agree with me on the paradox of thrift, but today even old-style Keynesians accept the natural rate hypothesis, which says than demand doesn’t affect the long run average level of output, just the volatility. So even if higher saving did cause high unemployment, it would have no bearing on long run decisions over what sort of tax regime to implement, which affect the level of saving, not the volatility.
3. Now let’s say I’m wrong about both the paradox of thrift and the natural rate hypothesis. Suppose that we are permanently at the zero bound and the central bank is too conservative to push unemployment all the way down to the natural rate, but instead targets an unemployment rate that is 2% above normal. (Put aside the question of why wages and prices don’t eventually adjust. Let’s suppose there is some sort of “hysteresis” that keeps the unemployment rate fluctuating around a trend line 2% above normal.) In other words, I’ll take the most extreme Keynesian assumption I can think of. What then? Even in that case a consumption tax is optimal. Even if we are not at full employment. That’s because what matters is not the level of saving but rather changes in the share of GDP that is saved. And in the long run those net out to zero.
The day we start making long run optimal tax regime decisions based on their implications for the business cycle is the day we become a banana republic.
Much better are Yglesias’s comments on the difficulty of distinguishing between consumption and investment. Of course to some extent those problems are just as severe for an income tax (think 3 martini lunches, corporate jets, etc.) FWIW, I’m actually pretty progressive on those questions. I favor treating education as consumption investment (no VAT) and business lunches and corporate jets as consumption.) BTW, when proposing egalitarian income redistribution programs, do progressives allow the perfect to be the enemy of the good? Obviously not, and I salute them for that attitude.
[The original version had a typo of consumption instead of investment for education]
I should comment on one of Roth’s statements:
This is not really revelatory; I know these economists understand the paradox of thrift.
Define “understand.”
But they ignore and eschew it in their real-good, barter-based mental economic models. I would suggest that the explanation for this error of composition is revealed by Scott’s words: “morally grotesque.” Moralistic beliefs about how individual humans should behave make it impossible for many economists to embrace an aggregate economic reality of which they are fully cognizant.
I do agree that beliefs about what is “morally grotesque” make it impossible for economists to “embrace an aggregate economic reality of which they are fully cognizant,” which is why I ignore all progressive analysis of income inequality.
More seriously, it’s not so much that I have moralistic beliefs about how people should behave, but rather how they should be treated by the state. I believe that people should not have to pay a higher tax rate simply because they prefer future consumption to current consumption.
PS. For you new readers, Yglesias was stretching the truth a bit when he said I favored having the central bank “guarantee” full employment. I favor NGDP targeting, which would (hopefully) minimize sub-optimal employment fluctuations.
Tags:
8. April 2013 at 12:15
PS. For you new readers, Yglesias was stretching the truth a bit when he said I favored having the central bank “guarantee” full employment. I favor NGDP targeting, which would (hopefully) minimize sub-optimal employment fluctuations.
Not ‘hopefully’, it does:
http://thefaintofheart.wordpress.com/2013/03/22/genie-in-a-bottle/
8. April 2013 at 12:50
Both Yglesias and Roth are just falling for the same old consumptionist fallacy, and all the corrolary implications of it (paradox of thrift, liquidity trap, etc).
Ultimately, while they may not consciously realize it, and would likely react strongly against it if it were explained in detail, the truth is that consumptionism is based on the moral claim that holds parasites (unproductive consumers) are a benefit to the host (productive consumers). That’s their belief deep down. They hate being told this.
Morality aside however, if we’re talking strictly about output and employment in line with consumer desires, attacking cash holding with government inflation and spending, is counter-productive, because the real problem isn’t lack of aggregate this or that, it’s lack of specific market sector this and too much specific market sector that.
8. April 2013 at 12:59
Why do people think that “demand doesn’t affect the long run average level of output”?
If demand fell to zero, output would fall to zero, right? There would be no reason for anyone to make anything, because nobody wanted anything.
8. April 2013 at 13:11
Prof Sumner, I promise you that I have looked for the answer to my next question on your blog, including reading your very excellent IMHO article for the general reader in National Affairs, “Retargeting the Fed.” Probably I missed the answer or it is so obvious I can’t see it. Here is my question: What I understand you to say that by acting more aggressively early when NGDP dips, the Fed can maintain full employment and avoid inflation. I assume that would be done by the usual open market actions, and then by maybe by QE if that fails. I get that. Head it off before we hit the zero bound and before NGDP crashes. But, right now, QE is limited primarily to government obligations, and in fact federal government obligations, like GSE paper. I suppose they could still buy up every piece of Treasury and GSE paper there is. But suppose that fails? At what point does the Fed stop asset purchases (assuming even that it has the legal power to go beyond the categories of assets it is buying now)? Corporate bonds? Munis? Stock market index funds? Apple shares? At some point, doesn’t this make you cringe in terms of the values of a free market economy? So I suppose my question is, now that we are at the zero bound, and assuming also that QE is as aggressive as it can be in terms of buying Federal obligations, what would you propose to do, and what are the limits? Let’s assume the helicopters are all grounded in a fog.
8. April 2013 at 13:15
“I favor treating education as consumption (no VAT) and business lunches and corporate jets as consumption.”
Educations as investment, surely?
The new version of the three martini lunch is the porcini-encrusted grass-fed beef at the company cafeteria, where Gourmet magazine raves that “over half the produce is organic.”
8. April 2013 at 13:17
Is there a good explanation on the delayed consumption argument for not taxing capital gains. I think in a world where people got growth-sized returns on investment it makes sense, but somehow I don’t think most investors in capital markets think of themselves as saving for later consumption (although that doesn’t make it true). Also I’m not sure how to prevent people from reallocating labor income as capital income.
8. April 2013 at 13:18
PS, I don’t think your FAQ’s really answer the question I posed above either — although I do gather you think QE has not been aggressive enough, and that is the problem. But let’s assume it’s not. There are limits to that too.
8. April 2013 at 13:41
“Saving is saving; it is defined in all the textbooks as the funds that go into investment.”
I don’t think this is true at all. Saving is defined in all of the textbooks as disposal income less consumption.
I have never seen a textbook that defines saving as funds that go into investment.
No, there are some algebraic manipulations of the supposed definition that income equals expenditure that imply that saving equals investment. But saving isn’t defined as investment.
Of course, the criticism really has to do with planned saving and planned investment. The argment from the definition that saving must be equal to investment because income equals expenditure is the paradox of thrift on steroids. An increase in planned saving results in less income and no change in saving which remains equal to whatever it is that firms were planning on spending on capital goods all along.
So, you tax consumption, and the result is that people plan to save more, but the result is that income is lower so that people save an unchanged amount, equal to whatever firms happened to want to spend on capital goods.
I don’t think this is right, and when people save more and purchase financial assets (and I don’t think finanical saving is meaningles,) this lowers interest rates, and makes purchasing capital goods more attractive to firms. How much firms spend on capital goods is coordinated with saving through changes in the interest rate.
Of course, if the central bank pegs the interest rate and fails to adjust it enough, this won’t work. And if the quantity of money is fixed and the demand for money is negatively related to the interest rate, it won’t work until prices and wages adjust enough so that the real quantity of money rises to match the added demand.
Saving is defined to be investment? To me this is bad and wrongheaded macroeconomics.
8. April 2013 at 14:20
Saying, “I favor treating education as consumption (no VAT [?]) and business lunches and corporate jets as consumption,” leaves unanswered indefinitely many issues about the dividing line between consumption and investment. Furthermore, it is obviously inaccurate, since some parts of education, and some business expenses for meals and travel, *really are* investment. You are close to saying: “I’d like to see taxes on consumption and not on investment, but in practice it’s too hard to distinguish between these, so I’ll just, by arbitrary rules, create a category X of all and only the items that are to be taxed, and *hope* that X overlaps substantially with consumption.”
(But I don’t want to sound a purely negative note. This post, like almost all your posts, is very insightful; that’s why I read them all!)
8. April 2013 at 14:25
Neil, True, but as long as NGDP fell to only 1 cent, then wages and prices would fall proportionately, maintaining full employment.
Maynard, You said;
“What I understand you to say that by acting more aggressively early when NGDP dips, the Fed can maintain full employment and avoid inflation. I assume that would be done by the usual open market actions, and then by maybe by QE if that fails. I get that.”
No, I’m saying the Fed should avoid causing NGDP crashes, not that they should “head them off.” The Fed caused this recession by its actions. It can avoid causing them with NGDPLT, regardless of whether we are at the zero bound or not. I don’t claim we would avoid all inflation, but it would be relatively low in the long run.
The scenario you contemplate would never occur in the real world with a policy of 5% NGDPLT. But if I’m wrong and it did then you could raise the target to 6%, or do negative IOR, or buy German and British government bonds. But again, it won’t happen because we simply aren’t that lucky. The US government cannot borrow unlimited amounts at zero rates if NGDP trends upward at 5% per year. Not happening. Unfortunately.
John, Thanks, I fixed that typo.
Bill, Every textbook I’ve seen has S=I as an identity. C+I = C+S.
Or add G and T and then you have government and private saving, which add up to national saving. And then add the international sector.
I’ll add more comments after dinner.
8. April 2013 at 14:38
“Saving is saving; it is defined in all the textbooks as the funds that go into investment.”
This is the central problem. Savings is not turned into investment; investment is turned into savings via a shift in the level of incomes. If investment falls and there is no rise in consumption, there is less money circulating. If the process continues over period of time, you get a situation where the total level of spending will fall and via that mechanism, incomes fall. This is the paradox of thrift.
If most textbooks assume that savings is turned into investment, then most textbooks are wrong. There is a difference between ex-post and ex-ante. Ex-post, savings equals investment, but ex-ante, savings doesn’t equal investment and the gap is filled with a financial sector that creates debt. This is why the financial sector exists.
Here is the primary paper where Keynes talks about this. He discusses this a bit in The General Theory, but he clarifies it in this paper:The Ex-Ante Theory of the Rate of Interest.
http://esepuba.files.wordpress.com/2011/10/keynes-the-ex-ante-theory-of-the-rate-of-interest.pdf
8. April 2013 at 14:40
Remember that if incomes are falling and people save at a certain rate, if you think about time being a factor, as time passes, the total amount saved would have been less than if more money would have been pumped in to fill the spending shortfall. What matters is the level of spending as the spending eventually becomes someone else’s income.
8. April 2013 at 15:08
Dr. Sumner:
A invests in and produces goods that he sets aside in inventory for now.
B is a consumer, and earns money, but does not purchase those goods in inventory, and instead accumulates money instead.
Who is the saver associated with the inventory, A or B?
8. April 2013 at 15:09
“What matters is the level of spending as the spending eventually becomes someone else’s income.”
Does the individual’s preference for money and goods “matter”?
8. April 2013 at 15:23
Bill, There is no “right or wrong,” it is strictly a question of semantics. I guess we used different textbooks. I’d add that some texts call disposable income minus consumption “private saving,” and T-G is public saving. But then it’s still true that total saving equals investment.
I think we both agree that planned saving doesn’t reduce income as long as the Fed targets NGDP or inflation.
Philo, The education comment was a typo–it should have been investment. I think as a matter of public policy it makes good sense to consider business lunches to be consumption. As for business jets, that’s a tougher call, but I’d consider them consumption as well. I do agree it’s a grey area.
8. April 2013 at 15:25
Suvy, I don’t find the ex post/ex ante distinction to be at all useful. To each their own.
8. April 2013 at 16:10
Scott,
“Saving is saving; it is defined in all the textbooks as the funds that go into investment.”
I agree with Bill Woolsey; this is a bad definition. Yes, S=I, but that’s different than saying that savings are “the funds that go into investment.” If I forego dinner at the cafe and put the money I would have spent under my mattress, then there’s no addition to investment (unless you count the lump in my mattress as investment). In fact, there may not even be an increase in aggregate S. The cafe owner’s income is less than it would have been had I bought dinner, so she might well save less. She may also reduce her expenditure, which will reduce the income, and, most likely, the saving of others.
It was precisely the fact that saving doesn’t go directly into investment that caught Keynes’s eye. The Classical economists dealt with a “corn economy” so that saving (not eating some corn) was directly related to investing (planting the seed corn that was saved). A modern monetary production economy is different, but, of course you know that.
The paradox of thrift pertains to the act of saving and the fallacy of composition. An individual can save more by reducing his expenditure, but if everyone tries this at the same time, you’ll get a much different result.
8. April 2013 at 16:16
Scott,
For PCT purposes. I think the important distinction is between business spending and consumer. Tax all consumer spending. Exempt all business spending.
The only modification I would make is to exempt expenditures in excess of $50k (or $100k) by consumers on durable goods, fixed assets, land, etc and instead impose instead on these items a state and local levy in the form of a fixed assets tax.
If you think three martini lunches and corporate jets are a problem, they are a governance problem not a tax problem. If they are not an efficient business expenditure, then shareholders presumably will limit or eliminate them.
As soon as you get into the weeds of what to tax and what not to tax, you give politicians huge unwanted powers and you put all the accountants and tax lawyers back to work instead of forcing them to get real jobs.
8. April 2013 at 16:16
Looks like MM is anti-market:
http://www.anirudhsethireport.com/blackrock-calls-for-bernanke-to-rein-in-qe-says-it-distorts-markets-risks-stoking-inflation/
“Fed policy has had a distorting effect on capital allocation decisions of all kinds at virtually every level of the economy.”
No wait, the market only consists of those managers who want more inflation.
Stupid me.
8. April 2013 at 17:33
Prof. Sumner,
I see that you wrote the book review below. Is there any place where we can access your review for free?
“The Missionary: McCloskey’s Apologia for Bourgeois Virtues and the Market” by Scott Sumner
http://onlinelibrary.wiley.com/doi/10.1111/j.1467-8594.2007.00310.x/full
8. April 2013 at 17:49
“Saving is saving; it is defined in all the textbooks as the funds that go into investment.”
If I took the difference between what I earn and what I spend and go to the bank and ask for green pieces of paper, which I hoard in a shoebox underneath my bed, these are not funds that go into investment, and this is not money that is saved. (by this definition)
If I put the difference between my earings and my consumption in a checking account that pays no interest, the bank would still be able to lend my money, and it could be directed to investment. So money in the bank is savings.
I am okay with the definition, but I want to make sure I am clear.
8. April 2013 at 18:32
Excellent blogging.
Yes, tax consumption.
BTW, I concur that savers are not sacrosanct, especially in relation to people risking capital in business start-ups or real estate development, or equity investors.
For generations we have heard sentiments that savers need to be protected or rewarded somehow by state actions.
Is the market trying to tell savers something now? LIke we don’t need any more savings, and that is why you get zero on your savings accounts?
8. April 2013 at 18:43
Doug M,
You said;
“If I took the difference between what I earn and what I spend and go to the bank and ask for green pieces of paper, which I hoard in a shoebox underneath my bed, these are not funds that go into investment, and this is not money that is saved. (by this definition)”
Except under MM, the Fed will offset this drop in velocity (NGDP) though increased OMP.
8. April 2013 at 19:21
“2. OK, most old-style Keynesians don’t agree with me on the paradox of thrift, but today even old-style Keynesians accept the natural rate hypothesis, which says than demand doesn’t affect the long run average level of output, just the volatility.”
They don’t. Larry Summers:
“Should we think of macroeconomics as being about–as it was thought about before Keynes, and came to be thought of again in the 1990s–cyclical fluctuations about a trend determined somewhere else, where the goal if you were successful was to reduce [the fluctuations’] amplitude; or as centrally about tragic accidents where millions of more people were unemployed for millions more person-years at costs of trillions of dollars in ways that were avoidable with more satisfactory economic arrangements.
Unless and until we adopt the second view, I think we are missing what is our principal opportunity to engage in human betterment.”
http://delong.typepad.com/sdj/2013/04/reconstructing-macroeconomics-exchange-mervyn-king-ben-bernanke-olivier-blanchard-axel-weber-larry-summers.html
8. April 2013 at 22:15
It is true by definition that on aggregate S=I (well actually S=I+NX), as aggregate savings is simply that which is produced but not consumed, countries simply can’t “save” by holding on to money.
The savings they are talking about is different, it is more the colloquial meaning of putting money in banks or stuffing it in mattresses(which appears to be the implicit assumption in many old Keynesian models that rely on MPC that for some reason still exist in modern textbooks) this under certain circumstances can lower the velocity of money, further depressing AD(that is how I interpret the paradox of thrift I’m not sure if that is an accurate view). I think it is clear that at least theoretically, if nominal interest rates are at 0, the paradox of thrift can happen, but it is the certainly the central banks fault if it does(as they should control AD). Economics is rife with examples of words being used in similar but very different ways, I think it was pretty clear they were not using the definition of savings from the part of the textbook dealing with GDP and growth, but instead the part that talks about how savings are a “leakage” in the circular flow. I am sure you are aware of all of this already, I am just wondering if I missed something.
9. April 2013 at 01:21
Bill: I strongly believe that Scott is right. Saving is direct result of investment. If I purchase a “capital good” (or whatever else as long as you can declare it as an “investment”) on January 1st from money in my wallet you can be sure that by the virtue of metaphysical necessity somebody somewhere in the economy has to “save” by the end of the year (or whatever period that you deem important, it can even be a period like 1 second). Like for instance I will buy tractor from tractor producer (and tractor is capital good). So at the moment of transaction tractor seller finds some cash on his account/in his wallet – which at this point become his savings (as money income that he has that was still not spent) If tractor seller never spends his money and he will still have them on his account comming December 31st, then S=I. If he spends it on consumption during the year, then it is just a musical chair game – tractor seller dissaves and somebody else will end up with the money as part of his income at which point they can be considered as “saved” and therefore it is still valid S=I. If he the tractor seller spends the money it on capital good – then by the definition this purchase is investment – and since it is also part of his income that is not spent on consumption it is automatically saving (or since it started in state of money “saved” we may say that the money were not dissaved from the point of view of tractor seller by action of purchase of capital goods)
It is not some miraculous mechanism, some magic fairy that equalizes savings with realized investment. It is a necessity stemming from the definition of the terms as they are used in economics.
9. April 2013 at 01:45
I must say I don’t understand the following statement:
“I believe that people should not have to pay a higher tax rate simply because they prefer future consumption to current consumption.”
It depends on the market rates. As long as net interest is higher than the inflation rate (which is the “normal” state of affairs) it’s not strictly true. You get more consumption (and so a lower tax rate) out of the same income if you wait.
You could make a much stronger argument for income volatility. Why should I pay more taxes if my income is very volatile? I am not sure it’s a big enough effect to really worry about.
I am not personally convinced by the assumptions that go into the models that prove the optimality of consumption taxation. I think the production function they use is too simplistic and that the results are mostly an artefact of that function. At some point capital could have zero productivity. On a macro level much “investment” has no productivity impact. Inventories of final goods (or raw materials) for example have no productivity implications. If I dig up more gold or iron or whatever and I lock it in a warehouse (which would count as investment), I will not get more output in the following periods. Only if I try to sell it, it will stimulate the economy by lowering prices…
9. April 2013 at 02:10
acarraro: Scott fights against double taxation. You first have your (wage) income taxed – so any part of income you save it was already taxed. But then you are taxed again when you receive interest from assets purchased from your already taxed savings.
This fact is independent from whatever inflation – or to be more precise – whatever real interest rate you accrue. Somebody may be thrifty even if facing with prospect of naturally diminishing savings.
For example if somebody prefers to can his food accepting the fact that it will depreciate in terms of taste and nutrition value – but which can be eaten all year long even in winter – then it is ones own decision. But why to tax someone thrifty just because other one marshmallow people ate all their food in summer when it was fresh and ripe and delicious? That was their decision so they should take responsibility for it.
It is not redistribution from less well-off to better well-of. It is plain and simple penalty for anyone who was born as two marshmallow person. But it is also more than that. It is discouraging thriftiness as some people who are 1 and half marshmallow when presented with this penalty may just eat it when available which brings that much more burden on two marshmallow people.
9. April 2013 at 02:17
Declan,
That’s too vague and indecisive to be an outright rejection of the natural rate hypothesis.
Education is an investment, but education and schooling are two different things: not all education takes place in schools and not everything that takes place in schools constitutes education.
9. April 2013 at 02:33
Because for every financial asset there is a corresponding liability, the net value for society as a whole is 0, and society-wide net inter-temporal transfer (saving)of financial assets is impossible.
This, of course, is not true for physical assets.
C and I are not consumption and investment, but expenditure on consumer goods and investment goods measured as added value. This means that I is spending on those durable goods that can be transferred to the future – savings. This ignores stockpiling of consumer goods.
Note that this saving is not investment in the sense of funding nor is it all spent on some mystic fungible substance called capital.
There’s still a question of how spending on investment goods maps to production capital.
Funding involves the transfer of financial assets and is thus outside the SNA framework used for GDP. Using GDP to reason about transfer of financial assets to fund production would seem a bit perverse.
GDP accounting was never designed for the uses many economists regularly make of it.
9. April 2013 at 02:53
Doug M
“If I took the difference between what I earn and what I spend and go to the bank and ask for green pieces of paper, which I hoard in a shoebox underneath my bed, these are not funds that go into investment, and this is not money that is saved. (by this definition)”
But wouldn’t one expect the central bank to offset people hoarding more currency by printing more of it thus increasing seigniorage the government receives (and similarly less money printing/seigniorage if people hoard less). So aren’t those green pieces of paper a lot like bonds, except that the government keeps all of the interest.
9. April 2013 at 03:23
“If I took the difference between what I earn and what I spend and go to the bank and ask for green pieces of paper, which I hoard in a shoebox underneath my bed, these are not funds that go into investment, and this is not money that is saved. (by this definition)”
Actually even case definition is perfectly in line with what Scott says – even without central bank Ossi mentions. If by the end of the year it is you who has green papers (that you earned as income during that year) in wallet and/or shoebox all it means is that somebody somewhere in an economy “invested” (aka purchased capital goods with her income) that year.
Because you could have earned money during a year only by receiving it from somebody else. And by the act of recieving money that somebody had to purchase something from you either for consumption or for investment – there just cannot be any other way. If he purchsed as investment (however you define it) then he was performing an act of “saving” and therefore your savings in form of green papers exactly reflect his investment. If he purchased something for consumption (however you define it) that automatically means that he had to lower his savings to exactly offset your newly gained savings in terms of green papers.
If in a hypothetical example there was no investment made in an economy during the whole year, then there was no savings created on net. As economy endowed some stock of green papers during start of the year the same green papers are there at the end of the year. It was just a process of one person dissaving funds she acrued previous year in order for another person to “save” (as not having enough time to spend money) as the year ends. Total net saving is constant.
PS: I highly recommend an excellent article by Andy Harless. He explains it much better than I can – http://blog.andyharless.com/2009/11/investment-makes-saving-possible.html
9. April 2013 at 03:28
Doug and all: Please sorry. I messed up with using the words “saving” and “savings” incorrectly in my previous post. While it makes sense to me it may not be that clear to somebody else. Just go read Andy’s article (link at the end of the post) he explains it all brilliantly.
9. April 2013 at 03:43
Saving isn’t defined as funds used for investment because like all good economic reasoning, it is built up from individual action. Saving is disposable income less consumption and implies an increase in net worth–accumulation of money, purchase of stocks or bonds, or real capital goods, or repayment of debts.
It is possible for saving to occur by investment, but not necessary.
Investment is the spending of a firm on some capital good. They do this with the plan of expanding production and revenue or reducing the use of other resouces, and so costs, and so add to profit in the future
The firms must finance this investment somehow. That is, they have to come up with funds to puchase the capital goods. This could be out of current revenues, out of retained earnings by selling stocks and bonds purchased in past, cashing in C.D.s purchased in the past, running down checking accounts built up in the past, selling new stocks or bonds, or obtaining bank loans.
Aggregate saving and aggregate investment are found by adding up all the individual amounts saved and all the individual individual amoutns invested (respectively.)
There is a claim that saving must be equal to investment as an implication of national income accounting. Accounting is about definitions and so the equality of saving and investment is matter of definitions.
I think this is true, but not interesting. And more importantly, it _is_ the paradox of thrift. In my view, the conventional wisdom is correct and Keynes was confused by definitions and equilibrium conditions. He balled them up in a confusd way. His followers straightened that part out (Hicks, for example.)
There is a market process by which an increase in _planned_ saving will lead to reduced expendenture and a lower equilibrium income and so reduced _planned_ saving so that it remains equal to an unchanged level of _planned_ investment.
And there are other market processes through which an increase in _planned_ saving by households results in purchases of various assets and lower market interest rates which provides an incentive for firms to increase their _planned_ investmnet.
Which process occurs? Both. And the process by which an increase in _planned_ saving results lower equilibrium income is one where the saving occurs by an accummulation of money balances and the quantity of money remains unchanged. (It also can work by saving by repaying bank loans and a reduction in the quantity of money.)
It is really just a confused version of the fundamental proposition of monetary theory. Plans to hold more money, with a given quantity of money, cause a higher market interest rate, reduced real output, or a reduced price level, so that the amount of money people plan to hold matches the amount that exists.
Now, the amount of money people actually hold is always equal to the amount of money that exits. If the “demand” for money is the amount of money people are actually holding, then the demand for money always equals the existing quantity of money.
If the quantity of money rises 10,000%, when people accept it in payment, they are obviously choosing to accept it, and so the demand for money just rises to match the supply.
No. The demand for money is how much money people choose to hold.
If velocity is defined to be nominal income divided by the price level, then the quantity of money times velocity must equal nominal income. It is just a definition. If the quantity of money rises, then given nominal income, velocity falls as a matter of definitions.
However, if we are interested in planned or desired velocity, this is no longer a matter of definitions. Of course, the rate at which money passes through one’s cash balances is a bit hard to think about for an individual. Instead, if the reciprocal of velocity is considered, k, then that is the demand for money as a fraction of nominal income. The actual value of k is always M/Y (quantity of money divided by nominal income.) It’s just a definition. But if k is instead understood as the amount of money people desire to hold, plan to hold, relative to their nominal income, then it is not.
By definition, the amount of apples sold must equal the amount of apples purchased. This is true regardless of the price. Therefore, a “high” price cannot lead to a surplus and a “low” price cannot lead to a shortage. Right?
No, because quantity supplied isn’t the amount sold and quantity demanded isnt’ the amount purchased. They are both defined in terms of plans to sell and plans to buy. Or perhaps desired selling and buying are better.
Every element of economics works this way. Including saving and investment.
Sumner’s claim that saving is defined to be investment is exactly like saying that the demand for money is defined to be the quantity of money created by the Fed or the supply of apples is defined to be the amount of apples purchased.
NO NO NO.
9. April 2013 at 04:11
Earnest question with a quick answer. Why do you put hysteresis in quotes? It always seemed like a fairly reasonable, intuitive assumption to make about the long-run effect of unemployment on people. And why do high-level economists debate about things like the existence of the paradox of thrift and hysteresis? Aren’t these easy questions about micro-foundations that have been extensively studied?
These just seem like empirical questions that should be settled by this point.
9. April 2013 at 04:18
If velocity is defined as nominal income divided by the quantity of money (not the price level.)
9. April 2013 at 04:44
Scott needs to respond to this: http://marginalrevolution.com/marginalrevolution/2013/04/some-thoughts-on-recent-japanese-monetary-policy.html
Also, China update: http://www.ft.com/intl/cms/s/0/591224c8-a0ff-11e2-bae1-00144feabdc0.html#axzz2Py9KkG3R
9. April 2013 at 04:46
Bill: “Investment is the spending of a firm on some capital good.”
That is your definition. Or it may be a shortcut how investment is tracked by statisticians is the easiest method in current environment. But generally accepted definition of investment is that it is spending of anybody (be it firm or individual) on something that is not consumption. Period.
And generally I agree with your claim that saving and investment are codetermined. That means that somebody may decide to invest instead of consume based on the interest rate that is in turn determined by supply of loanable funds. Or it can be the other way around – that is the meaning of the word “codetermined”, isn’t it?
Basically this is valid for any nonlinear system. Like for instance in a real world a velocity of an object is codetermined by the force applied to move the object (thrust) and by the force of the drag. By newtonian physics if an object is in “equlibrium” (let’s define it in a way that it is a velocity of not accelerating object) it is always so that drag equals thrust.
So yeah, you can define “drag” as that part of the force applied on the moving nonaccelerating object that is not “thrust”. And vice versa. And while I agree that this “identity” approach this does not explain the overall velocity of an object, or how the object behaves if environment changes, or if an object is not in equilibrium (accelerating or decelerating) etc. – none of that was subject of the discussion.
It was just used as an explanation why for instance things such as “financial thrust” has no meaning in the discussion about moving objects even if it may have meaning in some other context. Or at least this is how I understood the “thrust” Scott’s argument.
9. April 2013 at 05:25
Scott,
If I read him correctly Steve R accuses you of believing that if I forgo consumption and invest $X in the stock market, then “investment” (purchases of capital goods) in the economy will rise by $X. He talks about the fallacy of composition.
I was surprised you didn’t disavow this view. You don’t believe anything so daft, do you?
9. April 2013 at 06:05
Greg Hill, If it doesn’t go into investment, it’s not saving. I’ve addressed the money hoarding issue in other posts. I understand that people don’t like the definition, but it used in all the textbooks I’ve seen.
Cash is either a real asset (medium of exchange) or a government liability. Money hoarding either leads to more of a real asset (investment) or is offset by government dissaving (gov. liability.)
And please, you don’t need to explain what the paradox of thrift is to me, I’ve been teaching econ101 for 30 years. If you have some good objection to the reasons I reject the POT, I’d love to hear them.
dtoh, You said;
“If you think three martini lunches and corporate jets are a problem, they are a governance problem not a tax problem. If they are not an efficient business expenditure, then shareholders presumably will limit or eliminate them.”
This can’t possibly be true. Suppose the business provided a company car, a company mansion, a company clothes budget as well. This is a perfect way to evade taxes (unless you tax them as consumption), and hence it can’t be assumed efficient as it lowers the cost of compensating employees. If you don’t tax these things even I would oppose your plan. It would have zero chance of being enacted. People would form “businesses” just to avoid tax. VAT on consumption by the poor, and no VAT on the consumption of CEOs??
A business lunch is obviously consumption, I don’t even see it as a close call.
BTW, if you go that route your estimates of how much revenue that can be raised by a VAT (already unrealistically high) will be completely unreliable.
Geoff, I agree that bad monetary policy distorts capital decisions.
TravisV, I’m not sure.
Doug, I am talking about aggregate saving and investment, not individual. Where do you get the money to put under the bed? From the bank? Are they dissaving? Also, see my answer to Greg above.
Rob, There is only one useful definition of saving in my book. In other areas of econ we talk very straightforwardly about what’s going on. We distinguish between shifts in the the demand curve and the quantity demanded. We don’t call both “demand”. It makes no sense to use the term “saving” for both a quantity saved and a propensity to save. There’s no equivalent to “leakages” in S&D theory, thank God!
acarraro, You are wrong and it’s not really even debatable. If you are right I guarantee you will win a Nobel Prize in economics for revolutionizing tax theory. Google my watermelon/blueberries post, or read a Steve Landsburg poost on taxes.
Bill, Again, the textbooks say it’s an identity. You are free to prefer another definition, but you can’t claim that textbooks don’t say it’s an identity, as they do. If you use disposable income then you have private saving, which excludes government saving. And of course an individual can save without investing, but then someone else is dissaving an equal amount. I’m talking about aggregate saving and investment. If I hold more cash, someone else holds less. If the total stock of cash goes up, then see my answer to Greg Hill.
See my comments above to others about comparisons to S&D.
Matt, I don’t see much empirical evidence. We went from a decade of very high unemployment in the 1930s to low unemployment during the war, to normal unemployment after the war. Yes, at an individual level it can matter, but individuals don’t live forever. And more importantly, eventually you run out of money if you don’t work, so people eventually take menial jobs out of desperation. It’s one thing for a theory to be true in principle at the individual level, another for it to be statistically significant.
Saturos, I’ve already addressed all those MR comments in other posts. I agree with some and disagree with some. Obviously I think markets are efficient, so I take stock market reactions much more seriously than Tyler does. i agree that monetary stimulus will not boost RGDP growth dramatically, but it will certainly help.
Regarding the China article, I agree that China has lots of problems, but don’t have strong views on their debt situation. I am not that concerned about inflation in China, which is surprisingly low.
9. April 2013 at 06:05
This is actually a pretty good post from Yglesias, all things considered.
If you don’t buy it, it doesn’t get produced.
This assumes there is some ideal ratio of consumption/investment that is, for whatever reason one imagines, out of whack. I think sometimes people forget this is why we have market prices.
Let’s imagine, for instance, that some sort of brain parasite (say one similar to toxoplasmosis, which causes mice to become suicidal) evolves and spreads rapidly through the human population, markedly shifting behavior toward consumption and away from investment. What happens? Well, obviously, the returns to investment suddenly get much larger.
It’s pretty easy to see the opposite is true too: overinvestment pushes consumer prices down and the result is more consumptions.
This is how things are supposed to work. It is folly and arrogance to assume we know what the “right” ratio is at any given point in time, as opposed to letting prices do their job.
If there is a glut of investment as claimed, then consumption will become more attractive and investment less.
9. April 2013 at 06:25
Scott,
Well obviously it has to be a legitimate business expense (there adequate rules in place for this already), but I think you get drowned in the weeds if you start to regulate business dining, etc. Also if you would rather have $100 in business lunches than $70 in cash you obviously haven’t been to enough three martini lunches.
9. April 2013 at 06:30
Scott, do you think the Paradox of Thrift is not true over all time spans and all monetary regimes? That is, the paradox of thrift might not be true over 50 year intervals but it might be true over 2 year intervals. And it might not be true under an inflation targeting or NGDP targeting Fed, but it might be true if the Fed doesn’t counter decreased demand over the short term.
9. April 2013 at 06:30
Oh…. and as a business owner I would rather pay for an expensive lunch and have my employees working over the lunch than have them taking a lunch break and slumming it at McDonalds. When I was the banking business, we used to get our lunches provided free if we ate on the trading desk. If we ate outside of the office we had to pay for it on our own. Same deal on transportation. Stay past 8pm and you got a free taxi or town car to take you home. The firm was a partnership and the partners were pretty savvy about how they spent their money.
9. April 2013 at 06:31
Dr. Sumner:
“Geoff, I agree that bad monetary policy distorts capital decisions.”
Do agree with the assessment of Blackrock, that more inflation now will distort the markets even more? That’s what they’re saying.
9. April 2013 at 06:33
Why doesn’t anyone ever complain about the lengths of Woolsey’s posts? The absence of such complaints is strong evidence that the complaints over my posts are not actually because of their lengths as is typically charged.
9. April 2013 at 06:37
Saturos. Scott doesn’t need to respond to the Marginal Revolution blog on Japan. Cowen needs to think about it a bit more, that’s all. Hinting that the Government-sponsored Cen Bank moves are just “noise” is really poor when the stock market’s gone up 50% and the currency down 25%. Obviously, it’s more than just “noise”. If it was the price of restaurant meals in his home town that had gone up 50% I’m sure Cowen would have the answer.
9. April 2013 at 06:38
Geoff,
It’s about conciseness not actual length. Do your readers a favor and edit out the non-essential. As Cicero once said, “If I’d had more time, I would have written you a shorter letter.”
9. April 2013 at 06:55
Macro as taught to first year students is still kind of a mess, I have never figured out how you can throw a market for loanable funds or a money multiplier calculation right next to something that is based on MPC mattering, it is inconsistent madness. I agree that the only useful definition of saving is S=I, I merely pointed out that many intro textbooks later go on to discuss “savings” in other contexts.
9. April 2013 at 07:37
Dubois:
Investment is spending on newly produced capital goods.
It is not spending on anything other than consumer goods. Where do you get this stuff?
Scott:
I will waste some time looking at textbooks.
We will see how they define saving.
We will find out how many glossaries say Saving – this is the same thing as investment, the amount firms spend on capital goods.
9. April 2013 at 08:53
[…] Sumner has some nice remarks on this column here, but this bit caught by […]
9. April 2013 at 08:57
Scott, I have some thoughts on this post as well as Yglesias’ argument. We both ultimately agree that a consumption tax would be better economically (and morally, but for different reasons) regardless of business cycles.
But I think that there are rare instances in which a capital tax might work, as well. I also don’t by Rawls idea that “since it imposes a levy according to how much a person takes out of the common store of goods and not according to how much he contributes.” There’s no value of “contribution” without consumption, so there’s nothing morally virtuous about taxing consumption (except for reasons of simplicity, it’s immoral that we need tax lawyers to help us with a required function of society).
Would be interested to hear your thoughts: http://ashokarao.com/2013/04/09/the-phantom-line-between-consumption-and-production/
9. April 2013 at 09:19
dtoh, If businessmen are not enjoying their $100 lunches then they ought to go to Applebees, or TGIF, or a little inexpensive Japanese restaurant. You can’t have it both ways. A discussion can occur over any lunch, it makes no sense to order an expensive one unless people derive value from it. Unless of course it’s a bribe, but in that case we should also discourage it.
And again, if you exempt business consumption then your plan has zero chance of being enacted. Recall that consumption taxes already suffer the problem of being (falsely) perceived as being regressive. This would make it much worse.
BTW, There is no extra government discretion with the VAT applied to business lunches, as VAT would apply to all restaurant meals. The government discretion comes in when the government must decide which meals are consumption and which are not.
Travis, See my new post.
Geoff, I presume you know that I think inflation is a meaningless variable, so obviously I don’t think more inflation would matter. Inflation is just a number that government bureaucrats pull out of thin air. NGDP growth matters, and I favor NGDPLT along a 5% trend line.
rob, Good comment.
9. April 2013 at 09:24
Ashok, I had trouble following that post because I don’t know how you define terms like “piles of cash” and “capital” and “sitting on.” I don’t think rich people typically hold lots of currency.
I do agree that complexity is the number one problem with our tax system.
9. April 2013 at 09:49
I’m talking about cash as liquid assets like t-bills etc. (primarily held by corporations, not individuals – distinction I should have made).
My definition of capital is iffier. Because saving equals investment, there’s generally not a sense in which it makes sense talking about taxing liquid assets but, as I said, I think doing so (above a very high level) would incentivize reemployment, or at least an investment *in less liquid assets*.
9. April 2013 at 09:58
I.e. a tax on safe liquid assets during times of ultra-low rates to incentivize investment in things like factories or R&D projects.
9. April 2013 at 10:06
Would you agree , though, that there’s no difference in contribution and consumption in the way Rawls argues?
9. April 2013 at 10:49
I think I mean “sitting” in the same spirit you do in your FAQ:
“The Fed should stop paying interest on excess reserves, and if necessary should put a small interest penalty on excess reserves. This would encourage banks to stop sitting on all the money that has been injected into the system.”
9. April 2013 at 11:11
@Bill Woolsey
Investment is used in a number of senses, and people often aren’t very clear about which one they are using or assume it doesn’t matter.
1)In a total society-wide sense saving is net creation of physical goods that are conveyed forward in time. In that sense you can’t net save financial goods.
2)GDP separates spending based on whether it’s for investment goods or consumption goods. You can define the former as saving, and in SNA value added terms all spending is in those 2 categories, so expenditure – consumption goods = saving = investment. This not the same as 1. Note the choice of which category goods go into can be arbitrary and arguable, and there are services and intangible durables to consider.
3) Or saving can be defined as making available of financial assets for use as capital. This can be tightened up various ways, but it isn’t the same as 1 or 2.
This seems to be pretty good about the subject:
http://www.boeckler.de/pdf/p_imk_wp_100_2012.pdf
9. April 2013 at 12:49
I was wrong:
Cowen and Tabarrok: p. 484:
“Saving is income that is not spend on consumption goods. Investment is the purchase of new capital, things like tools, machinery and factories….. (investment is no buying stocks…)… Okay, let’s see how savings are moblized and transofmed into investments….”
Gwartney, Stroup, Sobel, and MacPherson, p. 533.
“Saving is income not spent on current consumption. Investment and saving are closely linked. Saving refers to the nonconsumption of income, while investment refers to the use of unconsumed income to produce a capital resource.”
Mankiw 267:
“private saving – the income that households have left after paying for taxes and consumption.” (definition in the margin)
But… Mankiw goes into this after a section called “Some Important Identies,” where he uses algebra to find that S = I.
But my favorite is Krugman and Wells. There is no entry in the glossary for saving. But one for the saving-investment identity.
on page 258, they cover the Saving-Investment Spending Identity. Only after that is covered, on page 259, do they define private saving. It isn’t in bold or anything. It says “National saving is equal to the sum of private saving and the budget balance, where private saving is disposable income (income after taxes) minus consumption.”
All of this talk of saving, without defining it.
So, the high Keynesians, both Mankiw and Krugman, go with Sumner. They start with the national income accounting identy.
They do define saving as income less consumption, but clearly it is an after thought.
I think they are two different things because I define them on a micro basis. Saving for each household and investment for each firm. Then each is aggregated. They are two different things.
But I also realize that because one person’s income is another person’s expenditure, realized saving must equal realized investment in aggregate.
But that doesn’t mean they are really the same thing.
By the way, Krugman explains “why” the identity is true using the paradox of thrift. If planned saving is greater than planned investment, then realized saving is matched by unplanned inventory investment. And firms will cut production and cause incomes to fall until planned saving equals planned investment.
Believe it or not, it was Nick Rowe who pointed out to me that if we have an economy with no inventories, the result is stronger and more direct.
In my view, planned investment and planned saving are what counts. As I explained, the identity of realized saving and investment in aggregate is of no economic interest.
It is no more interesting than the identity between the amount purchased and the amount sold. What is interesting is quantity demanded and quantity supplied and the role of price in coordinating the two. Gee, is price irrelevant because the actual amount purchased is never or more less than the actual amount sold?
Similarly, what is interesting about saving and investment is the role of the interest rate in coordinating planned saving and planned investment.
And of course, how monetary disequilibrium can interfere.
9. April 2013 at 16:41
dtoh:
“It’s about conciseness not actual length.”
Yeah, as if Bill ONLY gives “essential” comments (not knocking you Bill, just trying to show the blatant double standard).
Dr. Sumner:
“Geoff, I presume you know that I think inflation is a meaningless variable, so obviously I don’t think more inflation would matter. Inflation is just a number that government bureaucrats pull out of thin air. NGDP growth matters, and I favor NGDPLT along a 5% trend line.”
Well, obviously since Blackrock is saying more QE will distort the market, and since more QE will get NGDP higher than it is now, then Blackrock is equivalently saying that a higher NGDP growth rate that what currently exists (which is what you want), will distort the market as well.
Who cares if they call it more QE or higher NGDP? They’re talking about the same thing (increase in the extent of OMOs).
Are you saying they are wrong for claiming that a higher NGDP than what currently exists will distort the market?
10. April 2013 at 06:00
[…] Roth (1, 2), Scott Sumner (1, 2, 3), Bill Woolsey, and Matt Yglesias have been debating questions of saving versus investment […]
10. April 2013 at 06:46
Ashok, I’m not the one to ask about Rawls, I’ve never found his economics to be very interesting.
Bill, Thanks for that info.
10. April 2013 at 13:58
You keep asserting that income taxes have the same problems of categorization as consumption taxes, but you’ve never supported that statement.
You require the state to separate what is consumption from what is investment. This is fairly reasonable for the average person, so the perfect/good thing holds.
Separating consumption and investment is nearly impossible for the extremely wealthy. Your system would be extremely regressive.
The same problem is not intrinsic to income taxes.
10. April 2013 at 22:01
Scott,
You said;
“If businessmen are not enjoying their $100 lunches then they ought to go to Applebees, or TGIF, or a little inexpensive Japanese restaurant. You can’t have it both ways. A discussion can occur over any lunch, it makes no sense to order an expensive one unless people derive value from it. Unless of course it’s a bribe, but in that case we should also discourage it.”
Scott, by that logic professors should not be allowed private offices and should instead be doing all their work at picnic benches installed in a common faculty room….. or maybe they could just sit on the ground and spread out their books and papers on the floor.
“And again, if you exempt business consumption then your plan has zero chance of being enacted. Recall that consumption taxes already suffer the problem of being (falsely) perceived as being regressive. This would make it much worse.”
1) That’s why it’s progressive. 2) If you tax businesses then you have to structure it as VAT and hire a whole army of accountants and bookkeepers. Much easier to make it simple and just have a sales tax on individual consumption. Almost the same as a VAT and much cheaper and easier to implement.
12. April 2013 at 09:01
dtoh,
I was just checking to see if Sumner had addressed my argument that consumption and investment are impossible to tell apart for the elite, and now find myself defending him on a separate argument, haha:
Sumner never said disallow VIP lunches- he just said tax it.
I’m interested to read your response to Sumner’s criticism regarding avoidance strategies.
12. April 2013 at 09:32
myb6,
Why is this restricted to the elite? Think education “as a means to an end” (investment) vs. education “as an end in and of itself” (consumption).
15. April 2013 at 07:59
Ashok,
Oh, there’s definitely some gray area. However, I’m not convinced the gray area is numerically significant relative to the obvious consumption.
For people with serious capital, the consumptive value (control/status/risk-mgmt, etc) of their “investments” can be very significant relative to taxable consumption.