Assignment of targets to instruments, stability, and Functional Finance

J.W. Mason and Arjun Jayadev have a paper making a new (to me) point about the assignment of targets to instruments.

First I'm going to present an (over-?) simplified version of their model, to explain the gist of it. [I think I've got the gist of it, but I'm not 100% sure, and I know my simplification isn't 100% accurate.] Then I'm going to attack it (because it's a good paper and so worth attacking).

There are two instruments: M and F. There are two target variables: A and D, with targets A* and D* respectively. Which instrument do you assign to which target?

Suppose the picture looks like this:

You want to set M and F to get to the point where the red and green lines cross, so you hit both targets at the same time.

What happens if the two instruments are set sequentially, with each instrument assigned to one of the two targets, taking the other instrument as given? (Like a sorta sequential repeated but myopic Nash game)?

If F is assigned to target A, and M is assigned to target D, then (if we start at some point away from where the red and green lines cross) we follow a counter-clockwise spiral that approaches the desired point. Like the blue spiral followed counter-clockwise.

But under the opposite assignment, where F is assigned to D, and M assigned to A, we follow a clockwise spiral that gets further and further away from the desired point. Like the blue spiral followed clockwise.

The first assignment leads to a stable equilibrium; the second assignment leads to an unstable equilibrium.

If I had drawn a different diagram, with the red line flatter and the green line steeper, I could have got the opposite result. It all depends on the relative slopes of the two lines. To say the same thing another way: to avoid instability you need to assign targets to instruments according to their comparative advantages (like the relative slopes of the PPFs in the textbook Ricardian trade model). Which could go either way, depending on the parameters of the model (which is what J.W. and Arjun explore in their paper).

Attack 1. Stackelberg. If a government tells the central bank to target inflation (or NGDP) then the government is ipso facto telling the central bank to offset any changes that fiscal policy might have on inflation (or NGDP). In other words, the government is telling the central bank to do what is needed to ensure that the fiscal policy multiplier is always zero. It would be self-contradictory for a government to play Nash (acting as if it expected the central bank would hold its instrument constant) when it has told the central bank to adjust its instrument in response to any shocks including fiscal policy. A government that tells its central bank to target inflation (or NGDP), and that expects the central bank to do so, will instead be playing Stackelberg leader, picking a point on the central bank's reaction function. Under the standard assignment, if "A" represents Aggregate Demand, the government sets F by picking a point on the green line which represents the central bank's reaction function.

(Though, I gotta admit, sometimes I wonder if the current Canadian government and central bank totally get this point.)

Attack 2. Micro. There are always more targets than instruments. That's why economics is about trade-offs. Public Finance (which is what microeconomists call "fiscal policy" is as much about micro targets as macro targets. If the level of government spending, and the level of taxes, were both totally irrelevant, except for their effects on Aggregate Demand and long run fiscal sustainability (the things macroeconomists worry about) then it would be perfectly OK for macroeconomists to use fiscal policy for hitting macroeconomic targets, since microeconomists wouldn't care one way or the other. But microeconomists do care about the level of government spending and taxes, quite apart from their effects on Aggregate Demand. The whole point of assigning Aggregate Demand to monetary policy is so that microeconomists, and governments, can ignore Aggregate Demand when they worry about all the other trade-offs in public finance.

Take a very simplified view of micro public finance. The composition of spending between government and private spending matters. There are diminishing Marginal Benefits to government spending (for a given level of total spending), and that MB curve may shift over time (it shifts right in various emergencies, or when there's lots of kids needing new schools, etc.). And there are increasing Marginal Costs for tax revenue (because taxes are distorting, and the deadweight cost triangle rises with the square of the tax rate), and that MC curve may shift over time. So if the MB curve temporarily shifts right, with no shift in the MC curve, good microeconomic policy will be to run a temporary deficit, so the baby boom kids get new schools now but the taxes that pay for them are smoothed out over time for efficiency and equity.

The whole problem with the Functional Finance assignment of targets to instruments is that economists would say "Sorry kids, but you can't have the new schools you need, because the economy doesn't need any more Aggregate Demand just now". Or it would cause tax rates to predictably rise and fall over time in violation of the tax-smoothing motive of minimising the present value of deadweight losses (Jensen's Inequality).

Micro public finance matters. (And I need to write another post attacking that so-called "New View" of fiscal policy for totally ignoring micro public finance.)

Comments

[Market Fiscalist]

If either the CB or the fiscal authorities want to hit an inflation or NGDP target they can only (ultimately) do so by adjusting the money supply. In essence they do this (assuming its a positive target and V is fixed) by creating new money and either giving it away, or buying goods, services and assets with this new money. If they buy assets its called monetary policy, everything else is called fiscal policy (and may involve some additional steps where the fiscal authority initially sells bonds that the CB busy back).

Hitting the NGDP/inflation target is all about money creation and nothing to do with debt. Totally independently of this, the fiscal authorities can if they choose take out additional debt to fund the purchase of additional goods, services and assets. This will cause the debt/GDP ratios to rise. The only way that I can see that either the monetary or fiscal authorities could change the debt/gdp ratio is by changing (or ignoring) the inflation or NGDP target and varying the money supply in order to control NGDP and change its relationship to the nominal level of govt debt. In fact whenever they vary the inflation or NGDP target it will cause the current level of debt to change when measured as a proportion of NGDP.

So the variables are :
- Create money to buy stuff
- Borrow money to buy stuff

It doesn't matter who does it - these 2 things will determine both NGDP and GDP/debt ratios.

[Market Fiscalist]

If either the CB or the fiscal authorities want to hit an inflation or NGDP target they can only (ultimately) do so by adjusting the money supply. In essence they do this (assuming its a positive target and V is fixed) by creating new money and either giving it away, or buying goods, services and assets with this new money. If they buy assets its called monetary policy, everything else is called fiscal policy (and may involve some additional steps where the fiscal authority initially sells bonds that the CB busy back).

Hitting the NGDP/inflation target is all about money creation and nothing to do with debt. Totally independently of this, the fiscal authorities can if they choose take out additional debt to fund the purchase of additional goods, services and assets. This will cause the debt/GDP ratios to rise. The only way that I can see that either the monetary or fiscal authorities could change the debt/gdp ratio is by changing (or ignoring) the inflation or NGDP target and varying the money supply in order to control NGDP and change its relationship to the nominal level of govt debt. In fact whenever they vary the inflation or NGDP target it will cause the current level of debt to change when measured as a proportion of NGDP.

So the variables are :
- Create money to buy stuff
- Borrow money to buy stuff

It doesn't matter who does it - these 2 things will determine both NGDP and GDP/debt ratios.

[Ralph Musgrave]

“The whole point of assigning Aggregate Demand to monetary policy is so that microeconomists, and governments, can ignore Aggregate Demand when they worry about all the other trade-offs in public finance.” Two flaws in that argument.

First, it assumes that “microeconomists and governments” can’t think about more than one thing at once, which is a bit of an insult. Anyone who has got up to 101 in economics will have discovered that microeconomists often use equations with not one, but two, four, six or more variables.

Second, it’s very easy to have fiscal policy influence AD while NOT REQUIRING those administering fiscal details (e.g. schools) to think about macro matters. All you do is have the central bank (or some other committee of economists) decide how big the deficit shall be, while democratically elected politicians and voters decide how to allocate that money as between schools, defense, tax cuts etc. Indeed that split of responsibilities is entirely correct: technical matters rest with technicians, while obviously POLITICAL decisions (including what % of GDP goes to the public sector) remain with voters and politicians.

Indeed a system of exactly that sort was set out five years by Positive Money, Prof Richard Werner and the New Economics Foundation in their submission to the UK’s Independent Commission on Banking. See here (p.10-12 in particular).

http://b.3cdn.net/nefoundation/3a4f0c195967cb202b_p2m6beqpy.pdf

Incidentally that work advocates full reserve banking, but the existing or “fractional reserve” system is perfectly compatible with the above split of responsibilities.

[Ralph Musgrave]

One the subject of instruments and targets, I always think the grand daddy of “instrument and target” theory is the Nobel laureate economist Jan Tinbergen. He produced the so called “Tinbergen theory” which is something like “One policy instrument and one only should be assigned to each policy objective”.

[Vince Richardson]

I agree with Ralph. The best way to have fiscal stimulus influence aggregate demand is to have some sort of committee of economists decide the size of the deficit, while democratically elected politicians decide the obviously political stuff, like what proportion of GDP goes to the public sector.

[Nick Rowe]

MF: If the central bank sets M, and if V is fixed, then the red line is horizontal, so only one assignment is possible. But that is a special case.

Ralph (and Vince): do you understand what I mean by "Or it would cause tax rates to predictably rise and fall over time in violation of the tax-smoothing motive of minimising the present value of deadweight losses (Jensen's Inequality)."?

Because I suspect you don't.

[Luis Enrique]

Do you think that their point that the policy rate also influences the evolution of public debt is recognized by either academic economists or policy makers?

(I admit, I am not sure mainstream macro even tries to incorporate public finances into their models that try to tell us whether 'austerity' or stimulus is the right thing to do in a recession)

[Ralph Musgrave]

Nick,

I appreciate there are “losses” associated with failure to “smooth taxes”. I.e. there are costs associated with adjusting taxes. But likewise there are costs associated with adjusting the output of the capital goods industry (the sector of the economy that is supposed to adjust in reaction to interest rate changes).

In fact I’m 99% sure the difficulty / disruption / inefficiencies per dollar of adjustment will vary with the total size of the change (or more accurately with the percentage change in output required of whichever sectors/s of the economy are involved). Ergo if one million more people are to be employed, it makes a lot of sense to spread that increase over as wide a proportion of the economy as possible. The capital goods industry is not as large as the “government spending department” sector. In addition, adjusting fiscal stimulus can take the form of changes to tax, in which case the adjustment would be spread over almost the ENTIRE economy.

In short, if changes in output are all concentrated in the capital goods sector, that means far more dislocation for that sector (per employee or per dollar of output) than if the change is spread across all sectors.

And finally, there is a huge variety of skills among the unemployed. If they are to be moved into work as quickly as possible, there is clearly much to be said for them having as big a variety of jobs and types of employer to choose from as possible.

[Majromax]

I think your argument has merit with your graph, where all of the responses are nice straight lines, but politics interferes.

Imagine, for example, that the central bank is thin-skinned. When interest rates are below 1%, the Wall Street Journal writes rude editorials about "punishing savers" and "asset-price bubbles," and the economists from the Concrete Steppes start claiming that monetary policy is failing.

In this environment, there's an implicit extra target – the central bank wants to keep AD on track but it also wants to remain respectable in the locker rooms of finance. By calling for fiscal stimulus, the central bank is promising to not fully offset it and in so doing remain "conventional."

(From a mathie standpoint: choices of M and F lead to a two-dimensional coordinate space, where (A, D) = f(M,F); taking the partial derivatives of f with respect to A and D gives a 2 × 2 matrix, which when inverted gives the basis vectors necessary to change (locally) A or D alone. There is no choice about which instrument to assign which target; both instruments are used simultaneously and with different weights to correct for deviations from each target.)

[C Trombley]

Nick: Also posted this on Twitter, but wanted to make sure that people in the comments section saw that the comparative advantage & target stability thing was introduced in this paper by Mundell:

https://www.jstor.org/stable/3866082

[Market Fiscalist]

'The whole problem with the Functional Finance assignment of targets to instruments is that economists would say "Sorry kids, but you can't have the new schools you need, because the economy doesn't need any more Aggregate Demand just now"."

I'm not fully getting this. The govt could in theory decide how much spending it wants to do to meet its political and societal goals , then fund this via a combination of taxation, bond finance, and new money creation. The qty of new money creation would be the variable that is used to directly manage AD. It could then use bond finance as a buffer to prevent taxation jumping around too much as the govt deficit varied with AD management requirements. Having a sustainable debt/GDP ratio and level of taxation would be some of the things it needs to factor in when deciding how much it wants to spend to meet its goals.

[Nick Rowe]

C Trombley: good find! Closely related, but a bit different (because it's about external balance as opposed to debt sustainability).

[Scott Sumner]

I'm also puzzled that so few governments and central banks seem to get the monetary offset issue. You have Janet Yellen recently suggesting that more government spending would boost "demand", even as the Fed plans to raise rates in December to insure that "demand" doesn't rise, and cause them to overshoot their 2% inflation target.

This isn't rocket science. Are they missing something, or am I?

[JW Mason]

Hi Nick.

Thanks for the response. I do have to take issue with one thing tho. These arguments are not entirely new to you, since you wrote a nice post a couple of years ago in response to an earlier version of this argument.

On the substance. I don't think your first argument is quite right. The fact that the (inflation-targeting) central bank will offset any effects on output from the fiscal balance doesn't in any way contradict our argument, quite the opposite. The budget authorities don't care about the fiscal multiplier, in this story. They care about the debt ratio. The problem is that in targeting the debt ratio, they also change the interest rate the central bank needs to set to hit its target.

Now, you might say that it is irrational for the budget authorities not to take this response of the central bank into consideration -- and also irrational for the central bank not to consider the effect a change in interest rates will have on the fiscal balance chosen by debt ratio-targeting budget authorities. In some ideal world, the fiscal balance and interest rate would be chosen jointly to hit both targets. But our assumption that the two sets of policymakers are myopically pursuing their own targets isn't arbitrary, it describes --- it seems to me -- both how policy often really works and how many people think it should work. We really do live in a world of independent central banks with price stability targets. The designers of the ECB went to considerable lengths to, in effect, prevent it from considering the debt ratio target. And these cross-effects are largely ignored in public discussions -- how many people realize that the rise in the US debt ratio in the 1980s owes more to higher real interest rates than to primary deficits? So I think our model of a policy process that does not take into account the reaction of the other instrument, captures an important dimension of reality.

On your second point -- of course there are more than two targets, and more than two instruments. Models always simplify. One reason for adopting this particular simplification is that it is a standard one in textbooks and forecasting models.

The whole problem with the Functional Finance assignment of targets to instruments is that economists would say "Sorry kids, but you can't have the new schools you need, because the economy doesn't need any more Aggregate Demand just now".

But this is in no way specific to functional finance! The exact same issue arises if you have a conventional fiscal authority targeting the debt ratio. The only alternative is to keep building new schools until the marginal benefit drops to zero. There's no difference from the kids' point of view if the schools aren't being funded because aggregate demand is already too high, or because debt is already too high.

Of course you are right that when we get into concrete debates we can't just say "fiscal balance" we have to talk about concrete forms of revenue or spending. Presumably you don't want to adjust current spending on public services to manage aggregate demand, you want to use either broad transfers or taxes, or spending on infrastructure or other long-lived public goods where the benefits are spread out over many years but the demand effects on concentrated in the year they are purchased. If you say, "kids, we are not going to replace your aging but functional school building this year because all the construction workers already have jobs, we will wait for a year where a lot of them are unemployed" -- that doesn't sound so bad, does it? But this is a different level of abstraction than the main argument.

[JW Mason]

If either the CB or the fiscal authorities want to hit an inflation or NGDP target they can only (ultimately) do so by adjusting the money supply.

This is a background debate that the paper doesn't address. What is the mechanism by which the fiscal balance affects output? There's a point of view which says that it is only by changing the stock of government liabilities. That's Market Fiscalist here obviously, also old-school monetarism, and some versions of MMT. If you take that point of view, then the debt ratio and price stability are not two separate targets, they are two names for the same target. I personally would not want to go that route. Among other things, it suggests that deflation should be expansionary via some version of the Pigou or real balances effect. I don't think the people who emphasize government debt as net wealth for the private sector want to reach that conclusion, but it seems hard to avoid.

If we think of demand effects as purely about changes in the stock of money or other government liabilities, we also lose the ability to talk about different multipliers for different categories of spending and taxes, which is a problem when we come back to concrete policy questions. Personally I think it is better to think of the demand effects of the flow of fiscal surpluses/deficits as something distinct from the change in the stock of government debt.

I always think the grand daddy of “instrument and target” theory is the Nobel laureate economist Jan Tinbergen.

Met too! That's why he's in the subtitle of the paper. :-)

Are they missing something, or am I?

I think one thing you may be missing is the possibility that Yellen and others for whatever reason prefer a given level of demand with higher rates, to the same level with lower rates.

[Neil Wilson]

"If you say, "kids, we are not going to replace your aging but functional school building this year because all the construction workers already have jobs, we will wait for a year where a lot of them are unemployed""

This is precisely the problem with the mainstream analysis of functional finance. Because you are always talking about money and not stuff you are forgetting what the purpose of government is. And that then leads to the creation and attacking of Strawmen. The purpose of government is to take resources and deploy them for the public purpose. It is about what the government buys, not what it spends. The spending is a residual of the buying process.

So the actual decisions making goes like this. "We need to build a new school and we need these resources to do that. Now we have a look and see what is free and if there is nothing free *we stop some private sector activity from happening right now to make them free*". And we do it by some other mechanism than price competition.

That then moves onto the purpose of taxation. It has nothing to do with balancing anything - after all if government spends then you automatically generate the equivalent amount of taxation for any positive tax rate via the multiplier *unless people in the spend chain decide to save rather than spend* (which is what a deficit actually is). Taxation is actually there to stop people using stuff so government can use it instead. It is a way of altering the temporal order of activity and it is a really lousy tool for doing so. It's like carpet bombing when you have precision laser-guided weaponry. You are much better using the planning system, or the banking system to alter the order of things happening where possible.

In the school case you could suspend planning permission in an area until you get your offer for a school build accepted. Or you prevent banks lending on alternative projects. Then there is no need to tax specifically at all because you altered the temporal order of activity using more effective tools. (You prevent the banks creating money in a time/space period to allow government to create money in that time/space period instead). You'll note here that the government can always gets its price, because it can create a local recession until it does get its price. That's the power of a public monopoly over the currency

But let's say for general expenditure you'd probably stick with tax. And that leads to this description of how functional finance would work in practice. The currency issuer government gets first selection from the nation's resources to achieve the public good and it tends to match this with taxation/planning/banking policy because it always needs to happen regardless of the cycle. The private sector (and non-currency issuing governments/states) get to play with what is left via the usual bank lending spending chain mechanisms. Then the currency issuer government steps in and hoovers up any spare resources to create the nice to have public good. It can do this essentially for free. This is what the Job Guarantee enhanced auto-stabiliser system is there for. It handles the cyclical issues arising from the variable use by the private sector of free resources and the structural abandonment of certain resources (the unrequired unemployed labour).

You would never use required infrastructure for cyclical management. That's a complete Strawman.

It's when you add functional finance to the Job Guarantee that you get MMT. It is an integral part of the proposal that addresses the cyclical *effective* demand issue (i.e. addressing both demand and supply). Otherwise it's like talking about the mechanism of a grandfather clock and failing to discuss the pendulum weights.

[Nick Rowe]

JW: " I do have to take issue with one thing tho. These arguments are not entirely new to you, since you wrote a nice post a couple of years ago in response to an earlier version of this argument."

Oh God! My brain really is going.

Busy day today. Will return later.

[C Trombley]

Mason:

Minor comment - I was looking over the model and I think the equation right before equation 4 is wrong: it needs to be divided by (1-\tau i + \gamma b). (Also, equation 4 itself has the wrong sign, but you fix that in the later equations.) This doesn't change the analysis except there is now a line of singularities that y changes sign across, so if you're trying to apply the Mundell ideas Rowe looks at above carefully you'll see that the regions don't look like they did in the 1962 paper (so that, for instance, a central bank can move the economy from y>0 to y<0 without hitting the y=0 line).

Done slowly (for my sake) - You have as an IS curve Y = a + b Y + c Y* . I simplified the notation somewhat to make the issue clearer. The output gap ratio is defined y = (Y-Y*)/Y* or Y = (y+1)Y* . That gives an IS curve of (y+1)Y* = a + b (y+1)Y* + c Y* or y = (a-Y*)/Y* + b (y+1) + c . You define z = (a-Y*)/Y*, so this simplifies to y = z + b (y+1) + c or y = (z+b+c)/(1-b) . But you have y = z + b + c in the equation right before (4), which is incorrect. This won't change equation 4, because that assumes y = 0, true iff 0 = z + b + c .

[JW Mason]

You would never use required infrastructure for cyclical management. That's a complete Strawman.

And that Strawman is named ... John Maynard Keynes.

In 1930 he proposes

"to say to local authorities `If you will anticipate your expenditure, or produce anything that is at all reasonable we will let you have money out the Local Loan Funds below the standard rate of 5.25 per cent, or whatever it is now, going down to 4 per cent or even 3 per cent as a temporary expedient to employed only so long as there is a serious difficulty in home investment.'"

Which my line is a predict direct paraphrase of, tho he is thinking there more of tranportation and housing rather than schools.

Then in 1933 he publishes "The Means to Prosperity," which was the first major policy application of the multipleir and focuses on "schemes of capital development at home as a means to restore prosperity." From then on he almost always focused on public investment spending as the preferred tool for demand management. So in 1942 he writes, "I should not aim at attempting to compensate for cyclical fluctuations by means of the ordinary Budget. I should leave this duty to the capital Budget." Then a bit later he explicitly rejects the idea that taxes and transfers should be used for demand stabilization: "People have established standards of life. Nothing will upset them more than to be subject to pressure constantly to vary them up and down. A remission of taxation on which people could only rely for an indefinitely short period might have very limited effects in stimulating their consumption. ... it is better for all of us that periods of deficient expenditure should be made the occasion of capital development." And again (in a 1943 letter to Josiah Wedgewood), ""it is very much easier socially and politically to influence the rate of investment than to influence the rate of consumption." And to Wilfred Eady: "When we come down to strictly short-term fluctuations, such as those which arise out of the trade cycle, the alternative remedies are to try to off-set fluctuations of general demand by increasing investment, or to try to off-set it by stimulating consumption. Personally I favour the first alternative. ... capital expenditures contemplated by the authorities will be the essential balancing factor."

From the whole period from the 1930s to the end of his life. Keynes advocated shifting most investment decisions to the public sector. One reason was that he didn't believe that private enterprise would produce the level or direction of investment spending that was socially desirable, but the other reason was that he saw investment spending as the key tool of demand management. So in his proposal to separate the current and capital budgets of the British government, he writes, "the normal programme of the Public Capital Budget would have to aim at providing sufficient total investment to cover ... the current net savings of the private sector." And so on and so on.

There's a lot to agree with in MMT. But you guys are not the first people to think about these questions.

[JW Mason]

C Trombley - thanks.

[Jeff]

This whole discussion appears to be assuming one large fact that is not in evidence: that fiscal policy actually works. There is no empirical evidence I know of that it does, and many, many counterexamples. On the other hand, monetary policy has repeatedly demonstrated that it can affect NGDP in a somewhat predictable way, and since it can be adjusted very rapidly the prediction errors can be corrected for.

[Market Fiscalist]

JW,

You say "What is the mechanism by which the fiscal balance affects output? There's a point of view which says that it is only by changing the stock of government liabilities".

I would agree that fiscal balance can affect output but would make following points

- This is done by redistributing money in order to increase velocity. (Which is what happens when the govt sells bonds to savers, and then spends the money itself or gives ti people who will spend it).

- This will only work to address fluctuations in AD. I don't see how such redistribution of the existing money supply can be used to hit a NGDP or inflation target consistently over time . Eventually new money will be needed for this.

- If redistribution or infrastructure spending is needed then its needed whether or not AD is deficient any point in time. It therefor seems logical to me to use money to use creation/destruction to optimize AD, and to allow the govt to decide what level of spending it wants to undertake and whether this should be financed by borrowing or taxation.

[JW Mason]

If redistribution or infrastructure spending is needed then its needed whether or not AD is deficient any point in time.

In principle yes. But the thing about infrastructure is the benefits are spread out over time, while the demand effects are concentrated in the period it's built. So it's not crazy to think, as Keynes did, that you can move the spending up or back to smooth out fluctuations in aggregate demand. This is especially true if you think -- as he did, and I do -- that there is an immense backlist of socially useful public projects potentially available.

[rsj]

^ +1

I would point out that it's not true that infrastructure is needed regardless of AD deficiency or not, even in principle. That is like saying that

whether we purchase a good should be independent of the price of the good, we should only look at the benefits it provides

Infrastructure spending is used to build goods that yield a stream of benefits that must be discounted somehow, so for every project there is an interest rate at which you wouldn't undertake the project and another interest rate at which you would. Similarly there is a cost of materials and labor at which you wouldn't undertake it. These both capture different aspects of opportunity cost. Finally, if you believe that there are idle resources even if prices are sticky and market rates are too high, then still the opportunity cost of employing those resources is still lower than the cost of employing them if they needed to be bid away from the private sector.

So if you believe that there is such a thing as deficient AD, then by definition this is an environment in which the opportunity cost is lower than normal, whether due to rates, prices, or idle resources, which means that marginal projects make sense in that environment that wouldn't make in an non-deficient AD environment.

[JW Mason]

rsj: Right.

Nick: Your old post is here: http://worthwhile.typepad.com/worthwhile_canadian_initi/2014/02/principal-agent-and-the-assignment-of-targets-to-instruments.html

[Nick Rowe]

JW: Thanks. I had totally forgotten that old post! But it's quite different from your paper and this post. I think.

[GrkStav]

Neil Wilson says: "You would never use required infrastructure for cyclical management. That's a complete Strawman."

JW Mason responds with quotes from JM Keynes, as follows:

1. "to say to local authorities `If you will anticipate your expenditure, or produce anything that is at all reasonable we will let you have money out the Local Loan Funds below the standard rate of 5.25 per cent, or whatever it is now, going down to 4 per cent or even 3 per cent as a temporary expedient to [be]employed only so long as there is a serious difficulty in home investment.'"

I ask: Precisely what, in the above JM Keynes quote, indicates "**required infrastructure** is to be used for cyclical management"?

2."When we come down to strictly short-term fluctuations, such as those which arise out of the trade cycle, the alternative remedies are to try to off-set fluctuations of general demand by increasing investment, or to try to off-set it by stimulating consumption. Personally I favour the first alternative. ... capital expenditures contemplated by the authorities will be the essential balancing factor."

Again, I ask: What in the above quote by JM Keynes indicates "**required infrastructure** is to be used for cyclical management"?

I think the question is both 'semantic' and substantive. I think Neil Wilson has a "strong version" of "required infrastructure". The quotes from JM Keynes do not appear to me to conclusively gainsay his point. YMMV.

[GrkStav]

Jeff writes:

"This whole discussion appears to be assuming one large fact that is not in evidence: that fiscal policy actually works. There is no empirical evidence I know of that it does, and many, many counterexamples."

That "fiscal policy actually works" to do/effect what? Of fiscal policy having the objective effect of accomplishing *what* is there is an absence of evidence?

[JW Mason]

Precisely what, in the above JM Keynes quote, indicates "**required infrastructure** is to be used for cyclical management"?

Public capital expenditure means infrastructure. I don't think "required infrastructure" is a meaningful category, for the reasons given by rsj.

[Ralph Musgrave]

Scott Sumner asks why no one understands his pet theory, which he says consists of the idea that assuming the economy is near capacity, any fiscal stimulus will be offset by the central bank. Well I think we all understand that point.

Unfortunately Sumner normally takes his offset theory much further than that, and claims as far as I can see that fiscal stimulus is almost ALWAYS pointless because of “offset”. E.g. see his paper here:

https://www.mercatus.org/publication/why-fiscal-multiplier-roughly-zero-0

In contrast to where the economy is at or near capacity, when it is nowhere near capacity, the central bank probably WON’T offset. In the latter paper, Sumner is very vague on that distinction between when the central bank thinks the economy is at capacity, and when it thinks it isn't.

[Ralph Musgrave]

Jeff,

Where are these studies that show monetary policy works more quickly and predictably than fiscal? These two studies claim interest rate adjustments have little effect:

http://www.federalreserve.gov/pubs/feds/2014/201402/201402pap.pdf
http://www.usc.edu/schools/business/FBE/seminars/papers/ARF_3-29-13_KOTHARI-KLW_Invest.pdfAnd

I'm not the world's leading expert on this, but my IMPRESSION is that the lags and unpredictability are about the same for monetary and fiscal. Certainly the Bank of England claims the lag for monetary is about a year, which doesn't strike me as desperately quick.

J.W.Mason asks "What is the mechanism by which the fiscal balance affects output?"

Er...scratches head....the state prints money spends it on defense, education and health care. Result: employment among soldiers, teachers, nurses and doctors rises. Would that be it by any chance?
Alternatively, the state can borrow money (as Keynes pointed out) and do the same thing, but the effect is similar. Now what have I missed?

[Ralph Musgrave]

Grk Stav isn't sure what fiscal stimulus is supposed to do (scroll up about five comments). Would it be to raise numbers employed and GDP any chance?

[Jeff]

Ralph, I thought everybody agreed that that monetary policy both caused and ended the Great Depression and the inflation of the 1970's. And we have numerous examples of hyperinflations and debt-deflation episodes over the centuries, all caused by monetary policies. There just isn't any reasonable way to dispute this.

On the other hand, during many of the episodes cited above, fiscal policy was either orthogonal to or at loggerheads with monetary policy, and every time the monetary policy proved stronger. Reagan's tax cuts, for example, should have prevented the severe recession of the early 80's. The recession of 1974 should have slowed inflation. The postwar sharp drop in government spending should have caused another Depression after WW2. Scott Sumner has pointed out repeatedly that the economy picked up steam in 2013 during and following austerity.

What I'd like to hear from you is a single instance in which you can say that fiscal policy unamgiously "worked", i.e., that a deliberately increased deficit caused economic growth to rise, or that an intentionally increased surplus (or reduced deficit) had the opposite effect. To be convincing, you're going to have to show that monetary policy was not accomodating the fiscal policy, because I would take that as an indication that it was the monetary policy that had the effect, not the fiscal policy.

Really, to anyone with an even passing acquaintance with Friedman and Schwarz and more recent U.S. economic history, my contention should not be controversial.

[Ralph Musgrave]

Jeff, Thanks for your incisive questions.

Re your first paragraph, I’m not suggesting that monetary policy can’t have a big effect: as you say, various hyperinflation episodes prove it can. The crucial question is: which is better, monetary or fiscal stimulus, or should they be used in conjunction with each other?

As to why the sharp decline in government spending post WWII didn’t cause a depression, my explanation is that war spending was funded to a big extent out of borrowing. Then after the war (both in the UK and US) much of that borrowed money was paid back to the private sector. Thus the latter went on a spending spree.

Re your request for an example of where “fiscal policy unambiguously worked", I can’t provide one. But even if I could point to an example of where fiscal stimulus was followed by growth, that would not “unambiguously” prove a huge amount because of the amount of background noise. And that is of course a weakness in Sumner’s 2013 example.

Personally I object both to pure monetary policy and pure fiscal policy, i.e. I think the “in conjunction” option is best. The problem with pure monetary policy (i.e. interest rate adjustments) is that that involves interfering with the free market rate of interest and interfering with free markets reduces GDP all else equal. As to pure fiscal policy (i.e. government borrows $X and spends it), that’s defective because the effect of borrowing as such or in isolation is deflationary: the opposite of the intended effect. I.e. pure fiscal policy is like throwing dirt on your car before washing it.

Far as I can see, MMTers tend to favour the “conjunction” option: i.e. in a recession just have the state print money and spend it (and/or cut taxes). That was one option advocated by Keynes in the 1930s.

[Nick Rowe]

Thanks for all the comments. It's been a good discussion. I'm now feeling up to responding to some of them.

Scott: My guess is that maybe what they are really trying to say is "We want to use fiscal policy to raise interest rates, because we know central banks will offset it" but they can't say that politically. Instead, everything has to be translated into "more jobs!". I have noticed exactly the same translation of benefits into "more jobs!" for totally different policies, like free trade and improving the tax system, that have nothing to do with jobs (or aggregate demand).

JW: Thanks. I take it I got the gist of your article roughly right then?

"The designers of the ECB..."

Aha! There I think you might be onto something. Because I am assuming a single large fiscal authority, which can play Stackelberg Leader. But with many small fiscal authorities, that wouldn't work, and it makes a lot more sense to assume they play Nash. (I vaguely remember doing a couple of posts on the game between fiscal & monetary authorities, but can't remember if the Euro case came up.)

In general, I'm sympathetic to your idea of players groping (tatonnement) towards the equilibrium (Nash or not) of a game. (That's fundamentally why I dislike the Neo-Fisherian equilibrium.) But I *think* that groping would be more efficient at getting to equilibrium in a Stackelberg game than in a Nash game. (I must try to formalise that intuition a bit).

"But this is in no way specific to functional finance!"

I can get my head around the idea of a fiscal authority thinking about the trade-off between current micro desire vs long run fiscal sustainability. "We had better leave ourselves some fiscal room on debt/GDP ratio, in case we are faced with some future micro emergency". What I can't imagine is a central bank saying "There's a baby boom, so the kids need lotsa new schools, and the government only cares about AD, so we need to tighten monetary policy to get the government to build the new schools, not because the kids need them, but to create AD and jobs". It reminds me of that bit in Keynes GT, where he talks about monetary policy by the Trades Unions, only here it's education policy by the central bank.

Neil: "That then moves onto the purpose of taxation."

There are very good micro reasons for not using tax-finance to free up the resources needed to build the schools. Both equity (the kids themselves should pay for the benefits they get) and efficiency (raising tax rates now and then cutting them again when the schools have been built violates tax-smoothing and increases the deadweight losses of taxation) say the schools should (normally) be deficit-financed. Which means you need to use monetary policy to prevent AD from increasing.

Jeff: whether or not fiscal policy "works" to increase/decrease AD depends on what the central bank holds constant/targets (which is what's at issue here). If the CB targets AD, then fiscal policy can't work. But in general, if the CB targets something else (like a fixed exchange rate for example) then theory and evidence suggest fiscal policy will work.

[Jeff]

Nick, you say that "theory and evidence suggest fiscal policy will work." The theory part I find unconvincing, since I can just as easily theorize that it doesn't work. So I'd like to hear more about this evidence you refer to.

Your hypothetical that "if the CB targets something else (like a fixed exchange rate)" kind of implies that you think that, ceteris paribus, fiscal policy can affect the exchange rate. That may be true, but perhaps that's because people think the debt created by deficit spending will eventually be monetized. I don't know of a clear-cut example from history where fiscal policy was unambiguously effective. I do know that the benchmark case of Ricardian Equivalence says fiscal policy should not work, so if this is incorrect, I'd like to know when and where.

[Frank Restly]

Jeff,

Realize that central banks are a relatively new invention. The United States was formed in 1776, the U. S. Federal Reserve wasn't created until 1913. And so there is ample evidence that fiscal policy works in that fiscal / monetary policy were indistinguishable for almost 150 years.

But your point remains valid in that with separate fiscal and monetary authorities, it can be difficult to discern the individual effectiveness and accomplishments of either.

In Nick's example, I am presuming that M is the monetary policy instrument and F is the fiscal policy instrument. The assumption seems to be that M can be set at any value and place no constraints on the value that F can be set at and vice versa.

For instance, suppose M is an interest rate (set by the monetary authority), F is a deficit growth rate (set by the fiscal authority), the fiscal authority borrows to finance a deficit, and the fiscal authority is precluded from defaulting on its debts - can both M and F be set independently of each other? I don't think they can.

That isn't to say that there can't be instruments designed to operate totally independent of each other - but that would require persistent government budget neutrality or an alternative means of government finance.

[Jacques René Giguère]

Nick:"Which means you need to use monetary policy to prevent AD from increasing."
If we build more schools, construction workers and firms will move there instead of building shopping malls.Isn't allocating resources the role of the price system or has humble IO guy missed something during his micro courses?

[Jacques René Giguère]

Or in macro terms, there will be a crowding out of private investments. Which is exactly what you want. Why would the CB use monetary policy to prevent the price system from doing its job? Unless the CB wants to stop the building of schools. That is usurping the role of governments (which the paranoids (and THEY know who we are) have suspected for a long time...)

[Nick Rowe]

Jacques Rene: "Or in macro terms, there will be a crowding out of private investments. Which is exactly what you want."

Yes. (Plus a bit of extra saving/less consumption, which is also exactly what you want.) But that only happens if the central bank raises interest rates (or if you prefer, allows interest rates to rise) sufficiently to crowd out private investment (and consumption) 100%, so that AD stays the same, and stays on target.

[Jacques René Giguère]

Nick: aren't price rises sufficient to switch resources? Why do we need interest rae rise?
We don't ask the CB to intervene when Canadians switch from bathing suits to fur coats in early August (at least where I live).
Or is IO guy missing something real deep?

[Nick Rowe]

Jacques Rene: let there be schools built for government, and houses built for private buyers. We need builders to switch from building houses to schools, without creating an increased demand for the sum of the two. So we need the price of schools to go up, and the price of houses to go down, and we need the rate of interest to go up to bring the price of houses down.

You aren't missing anything deep. You are asking the right micro questions, to get a deeper understanding of macro.

[Jacques René Giguère]

Nick: are you telling me that you want to accelrate the process of shifting resources by crashing the massive and interest dependant housing in a way that you couldn,t do to the small non-interest sensitive bathing-suit-fur coats sector?
During WWI, when working at the Treasury managing the foreign exchange reserves, he met a friend who asked him his plans for the afternoon. "I am going to crash the spanish peseta!"

[Nick Rowe]

Jacques Rene: How would a Walrasian GE economy respond to the same shock of an increased government demand for schools? That's what we want a monetary exchange economy to duplicate.

[Jacques René Giguère]

If I understand you right, you indeed want to put gas on the fire and I'm ok with that.

[Ralph Musgrave]

Nick says that to get more schools built "we need the price of schools to go up". No we don't. If resources are shifted to school building and away from other stuff the demand for school buildings will probably raise the price of schools. But there is no "need" for that price rise.