« The Historical Statistics of Canada Are Back! | Main | Bank walks and options »

Comments

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

Usually monetary policy is done by buying government debt. Then the government debt and fiscal policy precedes monetary policy. The Central Bank could buy private asset but that is a step towards socialism. Which is, by the way, preferable?

So I also accuse monetarists of (implicitly) not caring about the size of government.

Jussi: having a government-owned central bank is a step towards socialism. If we had a purely private money+banking system, private banks would lend only to private firms and households. It is not obvious to me whether a government-owned central bank making private loans is more or less socialist.

When I consider "size of government", I think of the INFLUENCE of government. The influence of government is not the number of people employed by government.

I think monetary policy is aimed at decision makers, crafted to effect decisions by using the tool of money's physical availability and the tool of interest rates. These tools are more effective on the private sector than on the public sector.

I think fiscal policy is done by government. Government effects the larger economy by (for example) providing jobs, redistributing wealth, and creating rules for the economy to follow.

Roger: true, G/GDP is only one measure of the "size" of government.

But if we have a government-owned central bank, then we cannot talk about a world with *no* monetary policy. It doesn't make sense to say the central bank should "do nothing", because there are 1,001 different ways of doing "nothing". That's where "internet Austrians" so often go wrong (more sophisticated Austrians understand this problem, and want the government to get out of the money business altogether, or, as a second-best, try to mimic how they think a private monetary sector would behave).

I find Nick’s argument bizarre because it’s perfectly possible to increase fiscal stimulus while leaving total public spending unchanged: just cut taxes. That counts as fiscal policy doesn’t it? After all the definition of the word “fiscal” is something like “to do with taxes”.

You COULD COUNT tax cutting as monetary on the grounds that it leads to an increase in the money supply. Nevertheless I thought a tax cut was classified as fiscal.

"It is not obvious to me whether a government-owned central bank making private loans is more or less socialist."

Do you disagree that the use of the monetary policy implies change in the size of government? I guess you can argue that e.g. the ECB was loaning money but only against collateral (so the private bank legally retained the ownership). There seemed to be a limit for that. Now it makes purchases too.

This seems relevant:

http://worthwhile.typepad.com/worthwhile_canadian_initi/2011/10/the-optimum-size-of-the-central-bank.html

Jussi: cutting the inflation target from 4% to 2% would imply a bigger central bank. But whether we use fiscal or monetary policy to hit that target would have no effect on the mean size of the central bank, just the variance.

For a given inflation target, we may face a trade-off between variance in the size of G and variance in the size of the central bank. That would be a correct thing to say.

If I thought that government should do all the spending on education and nothing else and this spending should be fixed, and I also wanted to stabilize AD by varying the amount of base money then I could either:

1) buy assets for new money when I want to expand the money supply and sell assets for old money when I want to contract it

or

2) tax people less than the cost of education to increase the money supply and more when I want to reduce it.

I can see reasons to prefer 1 over 2, but I'm not getting why it necessarily means varying the size of government (unless govt size is defined purely by taxation).

I find Nick’s argument bizarre because it’s perfectly possible to increase fiscal stimulus while leaving total public spending unchanged: just cut taxes. That counts as fiscal policy doesn’t it? After all the word “fiscal” is defined as something like “to do with taxes”.

You COULD COUNT tax cutting as monetary on the grounds that it leads to an increase in the money supply. Nevertheless I thought a tax cut was classified as fiscal.

MF and Ralph: Distorting taxes have a marginal deadweight cost of tax revenue that is an increasing function of the tax rate. If I had written a longer post, I would also have talked about the increasing marginal (deadweight) costs of tax revenue. We need to look both at that increasing MC curve, and the decreasing MB curve, to get the optimal size of both G and T. Variance in T and variance in G, relative to their optima, are both bad things. But I wanted to keep it short, so just added that bit about "..adjusting for the marginal deadweight cost of taxes,.." to cover it.

I really do not directly relate "size of government" and the central bank. Perhaps this is a curious non-relationship.

Certainly the central banks and government are related. To confirm this relationship, I notice that anyone can print money. The trick is to LEGALLY print money. If we allow that "might makes right", then we have a fundamental relationship that connects government, central banks, and LEGAL money.

Central banks supposedly have control over the creation of LEGAL money. Of course, if governments control the central bank then government has control over the supply of money. Thus, the issue of the degree of government control over the central banks becomes an important political and economic issue.

Now return to the issue of size of government and the central banks. Remembering that I think NOT of government size, but of government INFLUENCE, then there comes forth a need to separate HOW government policy acts and HOW central bank policy acts. Government policy acts through spending and regulation. Central bank policy acts through monetary persuasions which is a much more psychological pursuit requiring evaluation of multiple prospective outcomes.

None of this fundamental background denies that government has the ability to control a central bank. Some will argue that "big government" implies a central bank controlled by government. It seems to me that "small government" has the same ability to control a central bank; the only difference is that small government MIGHT have made a choice to have a small influence on central bank polity.

"having a government-owned central bank is a step towards socialism"

No it isn't. Government is not the same thing as socialism.

"For a given inflation target, we may face a trade-off between variance in the size of G and variance in the size of the central bank. That would be a correct thing to say."

Agreed. So one needs to achieve a proper counter cyclical the change in size of the consolidated balance sheet (of the government plus central bank). But why is it more optimal to change the Central Bank's balance sheet than government's?

The central bank making private loans seem the best in theory (but ownership / collateralization line seems slippery slope)?

Nick,

Re your reference to “distorting taxes” are you saying that tax involves a cost in that it involves distorting what would otherwise be something resembling a perfect free market? That doesn’t strike me as a strong point because while a taxes obviously CAN HAVE a distortionary effect, they don’t have to. E.g. a sales tax on everything or almost everything that consumers buy is pretty distortion free.

Plus where there is an obvious distortion (e.g. high tax on the wealthy or on alcohol) that’s generally regarded as conferring net benefits. In which case it’s strange to refer to alleged costs of that distortion.

Ralph: Frances would give a better answer than me, but IIRC estimates of the marginal deadweight cost of tax revenue come in somewhere around 50 cents on the dollar (so $1.00 in tax revenue costs $1.50). (I'm probably wrong, but it's not trivial, and theory says it's an increasing function (because the area of a triangle is proportional to the square of the base, if height increases with base).

I'd imagine regardless of what the government is good or bad at providing, it still has a greater marginal product than the unemployment sector.

http://informationtransfereconomics.blogspot.com/2015/06/nick-rowe-shows-us-prior.html

(Unless it is infinitely bad!)

Jason: if monetary policy did not exist, that would matter.

"But if I really cared about the size of government, because I thought there were some things the government really should be buying, and other things the government really shouldn't be buying (initially high, but strongly diminishing marginal benefits of government spending at the optimal point) then I would strongly object to the use of fiscal policy instead of monetary policy to stabilise aggregate demand."

Does not follow. You ignore the contingency of fiscal policy.

"For example, if I thought that government should do all the spending on education, and nothing else, and if I thought there was an exactly right amount to spend on education, and that spending either too little or too much would be very wasteful, then I would strongly object to macroeconomists telling me to increase or decrease government spending to control aggregate demand. I would want a very strong covariance between government spending and the number of kids needing schools and teachers, and no covariance between government spending and the number of workers needing jobs."

Yes, but that is the result of your own stupidity. Of your belief that the gov't should only spend on education and nothing else. Then, of course, you could have too much or too little spending **within that confined focus**.

One virtue of fiscal policy is that it can be targeted. If it is a bad idea to increase school spending, then money can be spent elsewhere. Or, as Ralph Musgrave has pointed out, taxes could be reduced. They could be reduced in a targeted manner. Under certain peculiar circumstances taxes could even be increased on those who are hoarding money and decreased on those who will spend the money they would otherwise use to pay taxes.

"Monetary policy should handle that."

By comparison with fiscal policy, monetary policy is a blunt instrument. If there are too many workers needing jobs, can the central bank hire them? No, but the fiscal authority could, or subsidize their hiring. Can the central bank put money into the hands of poor people, who would spend it, thereby passing it along to others, such as local businesses, who could thereby retain their workforce or possibly increase it? Not in any country that I know of. But the fiscal authority could.

Monetary policy can respond quicker than fiscal policy, unless fiscal policy is automatic. But the events of the past seven years or so have not shown monetary policy to be particularly efficient or effective under current circumstances in many countries. Putting money into bank vaults does not necessarily get it into circulation. Yes, we avoided a second Great Depression in most countries. But in the US, for instance, that was largely the result of fiscal policy; i.e. massive bailouts. The problem there is that they were focused on Big Finance and Wall Street, with Main Street left wanting.

Nick, a question from a non-specialist. What about time. The gov't can spend more now (building roads and bridges is cheaper when the economy is bad) and spend less later. Surely it is at least theoretically possible to have a time-averaged fixed size of gov't and still use fiscal policy?

Chris: optimal government investment spending (just like optimal private investment spending) depends on things like the rate of interest, yes. So if the central bank lowers interest rates to prevent a recession, the government (like private firms) should spend more on investment. It's going to *look like* fiscal policy. But if it so happened that the government didn't have any investment projects that would meet the standard micro cost-benefit test at a lower interest rate, it shouldn't. (A lot of the extra investment spending the Canadian government did recently, like fixing bridges, could be justified in this way.)

Nick, I was thinking of cost of labour; I think the answer is the same, thanks. Surely though, the point of the fiscalists is people not working is guaranteed to be wasteful. Whereas people working, if less efficiently, still means net positive compared to the alternative (at the zlb.)

Chris: that would be the point of a fiscalist who forgets that monetary policy could also be used to increase demand, or who thinks it can't increase demand.

Bénassy neatly disposes of this issue, as does Atrios, by pointing out that fiscal policy can consist simply of giving people money, in amounts which vary countercyclically, and letting them decide how to spend it.

Philippe: "No it isn't. Government is not the same thing as socialism."

Sure, but I said: "having a government-owned central bank is a step towards socialism" because the government (or the central bank) that owns everything is socialism (by definition). And I think, like Nick I suppose, that a step towards will always have a cost. I'm not saying stimulus is bad at all times, not at all, but certainly the cost side has to be acknowledged too.

Ralph & Kevin: I don't get how the sales tax / giving out money for free is almost free of distortions. Again stretched to the limit this is a form of socialism isn't it?

But with from the discussion above this IMO seem to apply monetary policy just in the same way - it is no panacea and runs to similar problems at the margin.

Kevin: that would be like a negative income tax/GAI, financed by printing money. But it would have to be negative in some months. The poor wouldn't be happy if you told them their welfare cheques would be negative this month. Another way of thinking about this: in an optimal tax theory model, where the SWF is already maximised on the equality/efficiency trade-off, the shadow price of public funds (the lambda on the government budget constraint) will be positive, and a decreasing function of "outside" sources of public funds. (If you get what I am trying to say, not very clearly.)

Jussi: Kevin is right that lump sum transfer payments (which are just negative lump sum taxes) are free of distortions. Because the amount of money you get (or give) does not depend on your actions, so it has no disincentive effects at the margin. This is basic micro tax theory. But it doesn't work for sales taxes, because you can avoid them.

But Nick, Isn't that precisely the point? We are at the ZLB. Monetary policy is reduced to QE which is helpful but is also creating a bit of an stock bubble.

Nick: ah okay, I hastily lumped together Ralph's sales tax and giving out money for free. But yes they are different of course - lump sum transfers are as a first iteration distortion free. But is it plausible way?

How does it work if the government always just gives out money for free and presumable then finances itself by printing (I read your response to Kevin but he said "simply of giving people money, in amounts which vary countercyclically" which I take means no negative cheques)? So if it is inflation (tax) which gives up as a second iteration it surely comes with distortions?

I'm surprised no one has mentioned the obvious argument: it is reasonable to expect that one's assessment of the marginal net benefit of government spending would vary across the cycle. If the state of the world changes, why wouldn't your marginal assessments? To be specific, if the cycle results from herd behavior in the (private) investment sector, I could easily imagine altering my view of the optimal public/private investment mix according to the cycle.

That's the main point. Minor points: (1) Monetary policy through the asset market channel can be targeted, and is. QE in the US, for instance, targeted the residential housing sector. (2) The "publicness" of government spending can be tweaked simultaneously with activist fiscal policy: PPP's, privatizing old activities while initiating new ones, spinning off independent parastatals, etc. These are all part of the repertoire of modern governments. (3) In my view, the main reason conservatives have preferred monetary policy is that it can be "technocratized" and separated from the political sphere (and the risk of populism). Why else put such a premium on CB independence and focus so much research on identifying the optimal monetary policy rule? There is of course a vast literature on the merits and demerits of the two legs, but my sense is that monetary policy won out during the Great Moderation because, given your model, you could specify an optimal policy rule (usually), and it was politically "clean", not contaminated by democratic intrusions. (This also worked to establish policy credibility, which increased effectiveness.)

Personally, I think the people who are so sold on the deadweight loss of marginal tax rates should look to the international comparative literature, but that's a different topic altogether.

Nick & Jussi,
I can't see that the negative cheques problem should arise very often. It's quite normal for the stock of money to trend upwards at a moderate pace. More likely there would be some periods when people get smaller cheques than they are used to. Of course volatility like that may be especially troublesome for the poor, so an optimal setup would probably involve linking variance to income. When there is serious overheating you could have something like automatic VAT increases (admittedly distortionary) or interest-rate hikes.

But the main point here is that 'fiscalists' don't need to advocate silly projects, either in theory or in practice.

"But the main point here is that 'fiscalists' don't need to advocate silly projects, either in theory or in practice."

But can we say that "fiscalists'" projects are more non-optimal (i.e. less optimal) projects than those of "monetarists'"? I don't think so - but I'm not totally sure.

Yet I'm not totally sure which one is seen to offset which one (Nick once had post on this?) and why. Both seems to for instance suggests change in base money - at least in the long run. But are the changes similar/the same if the changes in consolidated balance sheet is the same. This probably varies from model to model but can we say anything more?

Peter: I think your main point is correct. And it's almost a semantic point, of whether we call that "fiscal policy" or not.

Trying to put some substance into that semantic question: suppose that government investment projects were decided by a bunch of microeconomists who were totally ignorant of macro, using only standard cost-benefit NPV > 0 sort of criteria, who thought (wrongly) that Keynesian macro was a load of weirdo macro BS, and that all multipliers were always zero, would they do this anyway? If the answer to that hypothetical is "yes", I say we should not call it "(macro) fiscal policy".

Nick: " Because the amount of money you get (or give) does not depend on your actions, so it has no disincentive effects at the margin." Not if my labor supply curve is backward bending.
Your income is not determined by your actions but your actions will be determined by your income. Gimme 10K and I won't teach that extra course. Gimme 100K and I'll stop teaching and comment all day (thinking about it, don't give me the 100K. My teaching is better than my comments.)

Jacques Rene: sure, but the marginal benefits of your actions will still equal your marginal costs. There's no tax wedge between the two margins. Standard demand, supply, tax wedge, and DWL triangle argument. A lump sum transfer just shifts the supply curve, it doesn't create any tax wedge at the margin.

"Socialism" is tribalism. "Owning" stuff is very Bourgeois in thinking. Thus the capitalist system owns everything anyway, including the government.

I don't think this analogy takes everything into account.

For one, government spending does not necessarily replace spending from household firms. It can be a transfer which is either spent, saved or invested by household firms. Automatic stabilizers in a recession like unemployment insurance are a form of transfer. Most of this money is spent.

So in a recession or prolonged slump, a government could boost aggregate demand with demand-side transfers. It could raise VAT rebate checks for the bottom 40%. Hand out temporary tax rebate checks for the 40% to 80% quintiles. And implement a countercyclical variable-rate VAT, which would be a consumption tax cut that encourages consumption. (In short, helicopter money for the bottom 80%.)

I think it's safe to say the supply-side, trickle-down, helicopter monetary-stimulus over the past 7 years has been a near-complete failure. Although it prevented a deflationary death spiral, it did not bring about a recovery in an adequate time frame. The goal of encouraging investors to invest and businesses to expand (creating the supply that hypothetically creates the demand) did not happen. Investors hoarded money and businesses held back because they had no confidence there was adequate aggregate demand to ensure their investments paid off. (Austerity acted as an anti-confidence measure.)

So if the goal of macro is to design a long-term "secular stagnation" then the New Classical and New Keynesian MMT proponents have done a bang up job. But if the goal is to have a high-growth economy that properly distributes incomes across all quintiles, I think it's obvious we have to return to a variant of the Keynesian system that accomplished exactly that in the post-war era (1945-1980.)

Ron: when I hear someone say "I think it's safe to say the supply-side, trickle-down, helicopter monetary-stimulus over the past 7 years has been..." I don't need them to complete the sentence before I know it is safe to say they they are talking complete rubbish. I'm not quite sure how to count the number of oxymorons in that string.

You sound like a bad parody of a high school debater on a soapbox, throwing together a string of cliches into a meaningless collage.

If I understood your arguments from a while back, Nick, if you're a New Keynesian then you can have fiscal stimulus just by cutting government spending... In the future. Promise to axe/cut lots lots of government spending programmes in several years time, and ride the FDR Express back to full employment.

Of course, to be THAT kind of New Keynesian, you'd actually have to be a fan of New Keynesian models, rather than a non-New Keynesian who uses New Keynesian models, and I found it interesting that there didn't seem to be many fans of New Keynesian models when you did those posts.

ChrisJ

"Monetary policy is reduced to QE which is helpful but is also creating a bit of an stock bubble."

I don't know whether or not current stock market values will or will not last, but I do think that the impression among central bankers (and those with influence over them) that QE has been doing this is an important reason why central bankers have not been bolder in these past 5 years or so.

W Peden: I think you understand my arguments correctly. (Though *that* sort of NK fiscal policy would still be a bad thing from the point of view of this post, because G would still have to vary over time, and so would sometimes be above or below the micro-optimal level of G.)

I'm very conflicted about NK macro. I've definitely got one foot in the NK camp. I attack NK macro more than I defend it, because it is something worth attacking, to try to fix up its weaker points.

On your response to Chris J: One argument for an increase in G that is not silly is that we might deliberately want to increase G in order to raise the natural rate of interest, because a low natural rate makes asset prices more volatile. We could argue about whether we call this "fiscal policy" in the macro sense.

Nick Rowe,

I may be straying off topic here, but is there a body of work explaining why volatile asset prices are a bad thing? If we're not reasoning from price changes, shouldn't the issue be WHY asset prices are volatile, rather than volatility as such? For example, if the price of houses has a sudden fall/rise due to a means of building cheap underwater houses/an earthquake, then volatile asset prices are an important part of optimal resource allocation.

Like Jussi, I wonder if the micro-optimality argument applies to the balance sheet/size of the base as well as it does to the size of government. Intuitively it doesn't, but I'd want to see an argument for an asymmetric treatment before I'm on board with your argument in this post.

W Peden: good question. I don't know. I was being sloppy in not distinguishing between "volatility" and "bubbliness".

There probably is some equivalent micro-optimality argument that could be applied to the variance of the monetary base. Variance in shoeleather costs would (presumably) be a bad thing. But I reckon we would be into second order of smalls here.

Nick: “And it's almost a semantic point, of whether we call that "fiscal policy" or not”

In my experience in business, the root causes of many disputes are often very basic issues like semantics. Also, it is often useful to set a challenge of trying to think of the simplest possible explanation for a difference of opinion. Here is a thought experiment.

Imagine two mainstream economists K and S who have broadly similar perspectives. Suppose that the only real difference is that K thinks that government bonds are money but S does not.

K will perceive that the money supply is increased when the government creates new bonds. He will also perceive that QE does not change the money supply as it replaces one type of money with another. QE is really just the central bank allowing the government to forego interest payments on its bonds.

S will perceive that the money supply is not increased when the government creates new bonds. However, he will perceive that QE does increase the money supply as it replaces non-money with money.

If both economists believe that, other things equal, an increase in the money supply will stimulate the economy then I’d suggest that K is likely to be a New Keynesian and S is likely to be a Market Monetarist.

So, are government bonds money?

Jamie: I was reading that, thinking K was more of a post-keynesian, or maybe MMT economist!

"So, are government bonds money?"

I read that some finance people use them as something pretty close to media of exchange. But even then they are beta money, which makes a difference.

Nick: “I was reading that, thinking K was more of a post-keynesian, or maybe MMT economist!”

That’s a fair point although I wasn’t intending to introduce PK / MMT into the debate. My intention was to find the simplest possible explanation of the difference of opinion between MM and NK economists. I guess I was also focusing on opinions at the zero lower bound although I didn’t say that in my comment.

There is a broader challenge to find the simplest possible explanation of the differences of opinion between the different sub-brands of PK or between PK and NK. That’s too difficult for this comment.

If you are correct then perhaps the simplest explanation is that PK and MM are pure / consistent / easy to understand in terms of their views of whether government bonds are money. NK is then best seen as some sort of hybrid which switches sides at the zero lower bound. NK is happy to be monetarist when interest rates are above zero and the central bank controls the economy by raising / lowering interest rates. However, it switches sides at the zero lower bound as it doesn’t really believe in QE or setting expectations.

Nick: “I read that some finance people use them as something pretty close to media of exchange”

Money can be a medium of exchange but it can also be a store of value. If government bonds are money then they are primarily a store of value. The simplest possible explanation of something often needs to exclude minor exceptions. In my experience, exceptions (and some finance people) are often the trees which prevent us from seeing the woods.

Nick: “But even then they are beta money, which makes a difference”

I agree that central bank money is different from other money. However, my distinction would be that government bonds are money with debtors attached (as are commercial bank loans). Hence, someone always has to use up some of their income to pay interest on this money, and someone else can act as what leftists call ‘rentiers’ and earn an income from this money. Central bank money doesn’t have a debtor attached so doesn’t earn rentiers an income. Hence, after selling bonds under QE, a pension fund has to try harder to earn an income compared with just sitting on its government bonds.

I missed it at the time: http://worthwhile.typepad.com/worthwhile_canadian_initi/2013/06/new-keynesian-countercyclical-fiscal-policy-isnt-what-you-probably-think-it-is.html

Nick: for me commenting is easier than teaching. Much less socially useful. Given that I reduced the socially useful, giving me free money does create a wedge. In the second degree but still a wedge.

"Ron: when I hear someone say 'I think it's safe to say the supply-side, trickle-down, helicopter monetary-stimulus over the past 7 years has been...' I don't need them to complete the sentence before I know it is safe to say they they are talking complete rubbish. I'm not quite sure how to count the number of oxymorons in that string."

Your definition of "oxymoron" must be different than the one I'm accustomed to (whose source is a dictionary.)

That particular sentence may have been contained clichés. But clichés are rarely oxymorons. The extraordinary monetary stimulus (zero-bound interest rates and Quantitative Easing) undertaken by the US Fed over the past 7 years is the very definition of Friedman helicopter money. It's certainly a supply-side form of stimulus. Ronnie Reagan used the phrase "trickle-down economics" to describe supply-side macro and also coined the slogan you referenced: "people spend money better than government."

But I think the premise of your hypothesis is wrong. If government uses fiscal stimulus in the form of temporary transfer payments and tax cuts it is not the government spending money; it's the people spending money however they choose. Of course, there are many examples of government spending money better than individuals privately, especially with public benefits and infrastructure.

What I really found odd about this blog entry was that an economist was using a hypothesis to attempt to disprove evidence: that is, attempting to prove monetary stimulus alone is the most efficient way to boost aggregate demand in a recession/slump when we are in the midst of a 7-year slump where this policy, put in practice, failed to work.

Scientists tend to let the evidence direct their models, not the other way around. Macro is certainly a very special form of science!

Ron: "It's [helicopter money] certainly a supply-side form of stimulus."

No.

Just to clarify, I'm not saying monetary stimulus is a part of supply-side ideology. I'm saying the stimulus is focused on the supply side of the economy. As Joseph Stiglitz puts it:

"Cheap money can lead to an investment boom in plants and equipment, strong growth, and sustained prosperity."

If one agrees that cheap money is a form of economic stimulus (there appear to be some ideologies opposed to this interpretation,) I believe one would have to agree this stimulus is mostly supply side. If any helicopter money winds up in the pockets of the average person boosting demand, it will have to percolate through the supply side of the economy first.

Ron: "If any helicopter money winds up in the pockets of the average person boosting demand, it will have to percolate through the supply side of the economy first."

No. Just stop.

"Ronnie Reagan used the phrase "trickle-down economics""

I'm giving this comment more attention than it deserves, but I would be amazed if someone could find a quote from Reagan using that phrase to describe his policies. It's one of those phrases that is used by some people as a substitute for thought, like many uses of the word "socialist" or "neo-liberal".

Jamie, are you going to be around for this post for awhile?

"Money can be a medium of exchange but it can also be a store of value. If government bonds are money then they are primarily a store of value. The simplest possible explanation of something often needs to exclude minor exceptions. In my experience, exceptions (and some finance people) are often the trees which prevent us from seeing the woods."

I would say government bonds are often used as means of clearing (MOC?), thus creating an alternative for the base money. In theory banks could clear in normal times without the reserves (assuming no requirements) but don't because they are the most convenient way. But in theory treasury could emit 1-day electronic t-bills to mimic base money. And store of value of course.

I like alpha-beta division and it is theoretically neat way. Yet in practice, IMO, the debt matters also. E.g. selling bonds is QE in reverse (consolidated) or that the changes in its present value affects the value of the currency.

"I agree that central bank money is different from other money. However, my distinction would be that government bonds are money with debtors attached (as are commercial bank loans). Hence, someone always has to use up some of their income to pay interest on this money, and someone else can act as what leftists call ‘rentiers’ and earn an income from this money. Central bank money doesn’t have a debtor attached so doesn’t earn rentiers an income. Hence, after selling bonds under QE, a pension fund has to try harder to earn an income compared with just sitting on its government bonds."

I would view base money and bonds having practically the same debtor behind them, the consolidated government balance sheet. The latter has to pay interest mostly because of the duration - and maybe secondly because the government might have a strategic option to default the debt - mainly through inflating it away. I would add that even zero is a number and in the end of the day what matters is the real return, also paying IOER blurs the picture.

W Peden: you are probably doing the right thing for the internet. Ron Waller has been commenting on economics blogs for a couple of years now. It is clear from his above comments he hasn't even learned the basics. If he had said "Despite what everyone else thinks, helicopter money is in fact a form of supply-side stimulus, not demand-side stimulus." he would be wrong but at least he would not be revealing his total failure to listen to and understand the conversation going on around him. By saying 'it is *certainly* supply-side stimulus' he does the latter.

Yes, "trickle down" is one of those phrases beloved by economically illiterate journalists and their readers, because it is a metaphor they understand. "Helicopter money" is a metaphor Milton Friedman used. Taken semi-literally, it describes money-financed transfer payments to a wide section of the population. Ron is just stringing together a bunch of cliches he doesn't understand into a self-contradictory melange. He is shouting slogans from his soapbox, waving his arms wildly, but not adding to the conversation.

There are other people, equally ignorant of economics, who actually do add to the conversation. At the very least, they don't subtract from it.

@Nick Rowe:

> Distorting taxes have a marginal deadweight cost of tax revenue that is an increasing function of the tax rate. If I had written a longer post, I would also have talked about the increasing marginal (deadweight) costs of tax revenue.

I don't think you could have written that longer post.

This post starts with the assumption (for the sake of argument) that the marginal benefits of C and G are always equal. This is inconsistent with an increasing marginal cost of tax revenue. By assuming such, you're making the assumption that the net utility of 100% G is lower (because of deadweight losses) than the utility of < 100% G.

If we assume that government spending only replaces private spending (that is, the government buys and distributes bananas rather than leaving people to use their own money to buy bananas), then we can still come up with a sensible toy model without increasing deadweight loss.

In regards to the broader tax discussion, I think we have to properly define "fiscal stimulus." I think that we can define a hierarchy:

* The narrowest definition is of balanced-budget fiscal policy, where increased government spending is financed immediately via tax revenues. In a strict old-Keynesian sense, this is stimulus because of marginal propensity to save/consume arguments. In a new-Keynesian sense, this is not stimulus because the net balance (G-T) matters.

* The next narrowest definition is of cyclically-balanced fiscal policy. Here, we expect Ricardian equivalence to hold, where a current deficit will be fully financed by a future surplus. This is probably the most common conception of "fiscal stimulus", but we run into an identification problem of what it means for the government to do nothing: is not cutting spending equivalent to stimulus? What of cuts to spending but further cuts to tax revenues? Is Greece actually one of the most stimulative countries under this definition, because of its large deficit?

* The broadest definition allows for a continual deficit, either through r < g arguments or through increases in the price level. Here, Ricardian equivalence is much, much less certain, being either found in the very long tail or buried behind inflation and seigniorage revenues. Inflation-financed deficit is of course only available to countries with an independent central bank, which means that this is probably a type of monetary policy even if it happens on the fiscal books.

Ron: I have unpublished your most recent 3 comments. Find a soapbox somewhere else.

Majro: (it's good to see a good comment!) "This post starts with the assumption (for the sake of argument) that the marginal benefits of C and G are always equal. This is inconsistent with an increasing marginal cost of tax revenue."

That's why I stuck in my clause about "adjusting for the marginal deadweight cost of taxes" in "The marginal benefits to government spending (adjusting for the marginal deadweight cost of taxes, and relative to the marginal benefits of private spending) did not diminish."

To be indifferent about the size of government, I would need to assume BOTH:

A. (Marginal benefits of G/marginal benefits of C+I) is independent of G/Y

AND

B. Marginal deadweight costs of T is independent of T/Y

(OR, just maybe, have A increasing at the same rate as B, which would be a really weird assumption.)

If you are saying that the hypothetical I needed to start the post with is a really really daft hypothetical...I would agree with you.

It's not a totally daft hypothetical, since it's true in a banana-economy. If people earn money by picking bananas and then use the money to buy bananas, then it doesn't matter whether they do so directly or whether the government taxes money, buys bananas on the peoples' behalf, and distributes the bananas.

That's the trivial reduction of assuming that the level of G is indifferent -- that the goods provided by G are identical to the goods provided by everything-but-G.

On the other hand, this stacks the deck in favour of old-Keynesian thought. Endogenous investment doesn't exist in this framework, therefore marginal propensities to save and consume take on exactly their "high school economics" significance. All unemployment is necessarily an AD deficiency.

On the gripping hand, *part* of government spending is an exact substitute for private spending: transfer payments. A government levying taxes (or taking on debt) to build a bridge is providing something distinct from the private sector, so we can talk about an optimal level maximizing society-wide utility for a particular utility function[1]. However, taxing me to subsidize you (or vice versa) has far more ambiguous effects. Borrowing from me to subsidize you further muddies the waters.

Ultimately, I think that borrowing for government transfers is an interesting component of aggregate demand. If there is no Ricardian equivalence, then the borrowing is a net AD increase as the demand curve for money shifts up. If there is full Ricardian equivalence, then the borrowing is fully and instantly compensated for by changes in private savings to pay back the expected future taxation and AD remains unchanged. Perhaps we can distinguish between these cases based on crowding-out and increases in interest rates?

[1] -- And going over or under this optimum leads to supply-side effects from a mis-allocation of real resources. That is fundamentally not a story about money, and it works if the Pharaoh is conscripting labour to build pyramids.

Majro: "It's not a totally daft hypothetical, since it's true in a banana-economy. If people earn money by picking bananas and then use the money to buy bananas, then it doesn't matter whether they do so directly or whether the government taxes money, buys bananas on the peoples' behalf, and distributes the bananas."

Yes, if the taxes were lump-sum taxes.

But yes, this does get into the question of Ricardian Equivalence. In a normal NK model, even when RE holds, increases in G still affect AD (for given r). But if government spending is a perfect substitute for private spending (the government does our shopping for us) then if the government buys everyone a banana we just all buy one banana less, eating the same total number of bananas, and even increases in G have no effect on AD.

> In a normal NK model, even when RE holds, increases in G still affect AD (for given r).

... and now we're back to "what does the monetary authority hold constant?" Or perhaps even the dual questions of "what does the monetary authority target," and "what does the monetary authority use as its instrument?"

* If the authority targets the interest rate, then r will be constant and your statement will fully hold.
* If the authority targets the price level and adjusts the monetary base, then r won't be constant.
* If the authority targets the price level and adjusts the interest rate, then r won't be constant but we'll call it monetary offset.

Does the use of fiscal policy as an economic tool result in immediate budget cuts or increases?

Majro: yep.

financewhiz: people normally talk about it that way. But in some models, it's expected future government spending and taxes that matter as much (maybe even more) than current spending or taxes. And we could have a fiscal policy that amounted to making commitments about the future rather than changing G or T today. But when that future time arrives, the government would have to change G or T as it had previously promised, otherwise its fiscal policy would lose credibility.

Too Much Fed: Jamie, are you going to be around for this post for awhile?

Hi. I hope you are well. As you can see I am still reading this thread but I don't currently have anything else to add to my previous comments.

Nick: in recession, holding on to money is paramount, at the expense of banana-eating. G can buy a banana and I will eat it gladly even if I still hold to my money. There is no RE in these circumstances. (And the shadow-price of babanas is nil anyway.)

Jamie said: "So, are government bonds money?"

I am going to agree with Scott Sumner on this one. Gov't bonds are not "money" because they do not have a fixed conversion rate.

"K will perceive that the money supply is increased when the government creates new bonds. He will also perceive that QE does not change the money supply as it replaces one type of money with another."

What are these "another" type(s) of "money"?

"S will perceive that the money supply is not increased when the government creates new bonds. However, he will perceive that QE does increase the money supply as it replaces non-money with money."

Let's assume a commercial bank does "QE". Does the "money" supply increase?

"Nick: “But even then they are beta money, which makes a difference”

I agree that central bank money is different from other money. However, my distinction would be that government bonds are money with debtors attached (as are commercial bank loans)."

I disagree that gov't bonds are "beta money" and disagree that demand deposits of the commercial banks are "beta money". I believe demand deposits of the commercial banks are both medium of account (MOA) and medium of exchange (MOE) because they have a fixed conversion rate (it just happens to be 1 to 1) thru the central bank to currency. That makes those demand deposits "alpha money".

From:

http://worthwhile.typepad.com/worthwhile_canadian_initi/2014/12/alpha-beta-and-gold.html

"csissoko: Yours is a fair point. When there's a run on a beta bank, the alpha leader will temporarily act like a beta. If there's a run on all the beta banks, the alpha might have to choose a different path, because the beta wolves can't keep up with him. The Bank of Canada would have to allow inflation to exceed its 2% target, to bail out the commercial banks.

If this happened all the time, and if betas were never disciplined for failing to follow the alpha and needing the alpha to bail them out, then my alpha-beta distinction would be meaningless."

Jamie, I am not sure if am reading that correctly, but I believe some version of it is true. If there are runs on all the "beta" commercial banks along with all of them being solvent, the central bank will "bail them out" all of the time by increasing currency so the alpha-beta distinction is meaningless.

Thoughts on those?

Hope you are well too!

Ok Nick I'll just put it out there. I'm presuming this post was inspired by a piece Scott did on fiscal policy and the size of government. The piece was literally called 'fiscal policy isn't about the size of government.'

http://www.themoneyillusion.com/?p=29763

Do you agree with the post-because he seemed to be saying it's not about the size of government while your post seems to say it is about the size of government.

In a way then are you saying that monetary policy is preferable assuming that government is already the optimal size prior to the need for stimulus?

Mike: put G/Y on the horizontal axis. (That's a measure of the size of government, as a fraction of GDP.) Draw a downward-sloping Marginal Benefit curve, and an upward-sloping Marginal Cost curve. Where (you think) those two curves cross determines (what you think is) the optimal size of government.

Scott is quite correct in saying that your beliefs about where the curves cross should have no effect on whether you agree or disagree with the use of fiscal policy.

I am saying that (your beliefs about) the slope of the two curves where they cross should affect how much you dislike fiscal policy. If (you think) both curves are steep, you will strongly dislike fiscal policy. If (you think) both curves are fairly flat, you won't mind fiscal policy much.

It's not the size, it's the variance around that size.

Why am I using a visual explanation for an artsie?

"There are other people, equally ignorant of economics, who actually do add to the conversation. At the very least, they don't subtract from it."

Yes... that last bit there: that's what I inspire to be! Lol

or maybe, "that's what I aspire to"

OK that was mildly brilliant.

Trying to pin them down on the idealk ratio is... interesting. Martin Wolf once told me he thought WW II was a good starting point(!).

Kevin: But the main point here is that 'fiscalists' don't need to advocate silly projects, either in theory or in practice.

Heh, sending people checks for nothing has somehow become less silly than a Bridge to Nowhere or a Hole-Digging Brigade.

That amounts to repealing welfare reform, EITC, and renouncing the entire philosophy behind them. Incentives matter.

Anyways, exchanging money for nothing is needlessly distortionary when the Fed can bid on every asset in existence.

TallDave: thanks.

It's also tricky in practice even to measure the size of government. If the government sends every individual a cheque for $1 million, and also imposes a tax on every individual of $1 million, it looks like government just got a lot bigger, but nothing has really changed, except for the admin costs. And if we measure it by government expenditure, as opposed to transfer payments, is there much difference between the government buying you glasses, and giving you the money to buy glasses? (Though maybe you get the style you want.)

Tom: you're doing fine.

Too Much Fed: “Thoughts on those?”

My focus in this thread has been the thought experiment I described in my first post. What I don’t want to do is to discuss whether specific non-economists like you or me think that government bonds are (or are not) money. I’m interested in how real economists think about this and whether their belief on the matter is a good predictor of whether they are New Keynesian (bonds are money) or Market Monetarist (bonds are not money).

With that in mind, you say: “I am going to agree with Scott Sumner on this one. Gov't bonds are not "money"”. If you are correct that Scott believes that government bonds are not money then that is consistent with my thought experiment. Note however that Nick said that he thought my thought experiment more accurately described the difference between Post Keynesians and Market Monetarists.

The key here is that minor differences in semantics can lead to major differences in policy prescriptions. I’m not sure that there is a “correct” definition of money. However, that may limit any ability to come to a consensus on policy matters.

> It's also tricky in practice even to measure the size of government.

I'm not sure it's so much "tricky" as "ill-defined."

Off the top of my head, I can think of two slightly different definitions of "big government."

The first is a "big government" in real terms, in that ultimately the government makes purchasing decisions that impact a significant chunk of the real economy. This is the type of government that builds bridges and bombs and employs police and park rangers. Indirectly, this is also the government that provides targeted subsidies (operating at the extrinsic margin, anyway), for housing and health-care.

This sort of "big government" works without money, in that in theory the government could just conscript the labour required to do its thing.

The second sort of "big government" only works with markets and money: it is the redistributively big government. Here, the government is largely indifferent to what is purchased, but it cares very much about who makes those purchases. This is the government of progressive taxation and cash transfer payments. It is essentially impossible to have this sort of government without money, since there is no conceivable set of a priori real transfers that would result in the same sort of change in purchasing power as a transfer of money.

We can then have differing notions about the proper size of government on both axes. Economic analysis (and on a micro-scale, cost-benefit analysis) helps much more with discussions about the real-terms size than the transfer-size.

Nick,

Actually, Milton Friedman devised a way to measure the size of the U. S. government by counting the number of entries into the federal register. Only, the U. S. government kept shrinking the size of the font used for printing the register.

Majormax,

In the spirit of the U. S. Constitution, there is consolidated versus divided government. With a consolidated government there are no political parties, no separate branches of government (judicial, executive, legislative), and no separate layers of government (local, state, federal).

There is a tradeoff between large democracies and small autocracies. With autocracies, there are fewer checks on the power that a government official can wield.

Jamie said: " I’m interested in how real economists think about this and whether their belief on the matter is a good predictor of whether they are New Keynesian (bonds are money) or Market Monetarist (bonds are not money)."

Do New Keynesians actually think bonds are "money"?

"I’m not sure that there is a “correct” definition of money. However, that may limit any ability to come to a consensus on policy matters."

Exactly on the consensus part.

I think you need 3 groups.

The first one thinks that the monetary base is "money" and that bonds are not "money".

The second one thinks that the monetary base is "money" and that bonds are "money".

The third one thinks that currency is "money" and that certain demand deposits are "money". Also, are there any "real economists" in the last group?

But if I really cared about the size of government, because I thought there were some things the government really should be buying, and other things the government really shouldn't be buying (initially high, but strongly diminishing marginal benefits of government spending at the optimal point) then I would strongly object to the use of fiscal policy instead of monetary policy to stabilise aggregate demand.

Why doesn't the same apply to the CB buying private assets?

On the one hand, you say that there is an optimal amount that government should spend on education that correlates strongly with the amount of pupils and teachers. Spending more (or otherwise increasing the amount of money that schools get to keep, say by reducing taxes?) increases the size of government with diminishing returns. Fine.

On the other hand, If I may misrepresent your position in my own words, you would say that if the CB bought shares in private schools this would leave them with more cash which they would then spend on more teachers or more school buildings or something similar. In any case, the schools could not just go out and spend the money on hookers and blow. The board wouldn't approve. Is that not an increase in the size of government? Considering they could only spend the money on things related to schooling, how is that not shifting the size of schooling beyond the optimal point? Government owns more school assets and has artificially increased the size of schooling relative to what the market was willing to fund before. Either that or you're conceding that monetary policy has no effect other than shifting ownership claims onto government books in exchange for money. Why aren't there diminishing marginal benefits to monetary policy? Because there aren't any benefits in the first place? I somehow doubt that is what you're saying.

What am I missing? Is it only about the inefficiency of taxation? That seems to me not about the size of goverment but about the optimal size of different sectors of the economy. Of course government might be bad at hitting that spot. But then so might CBs.

Also, where do you draw the line between fiscal and monetary policy? I've only ever seen you write about what monetary policy is and possibly what fiscal isn't, but not what fiscal policy is. Maybe you have a post somewhere along the lines of Majro's comment above I could read?

Actually, forget the above. I saw my mistake. The schools aren't the ones who own the assets. Sorry for wasting webspace...

The comments to this entry are closed.

Search this site

  • Google

    WWW
    worthwhile.typepad.com
Blog powered by Typepad