[Apologies for the title. It's a Canadian joke, about our fixations, that can be told various ways, but always ends with the punchline: "Elephants: a federal or provincial issue?"]
Does fiscal policy affect the price level? If so, are we talking about federal fiscal policy only, or are we talking about provincial fiscal policy too? What about municipal fiscal policy? And why stop at governments? Lots of corporations issue bonds. And individuals borrow too.
There are two differences between the federal and provincial (and municipal) governments:
1. The federal government gets the seigniorage profits from the Bank of Canada. If we assume a 5% currency/GDP ratio, and 5% nominal interest rates, (and ignore operating costs for simplicity) that's 0.25% of GDP. A nice little earner for the federal government, but no big deal. If the Bank of Canada gave its profits to the provinces instead, the provinces could cut tax rates a little, and the federal government could raise tax rates a little, but nothing much else would change.
2. The federal government can, though it would hesitate to use this power, order the Bank of Canada to stop targeting 2% inflation, and do something else instead.
We can ignore 1. It's simply a minor detail that belongs to the economics of fiscal federalism and transfer payments between different levels of government.
2 might matter. But as long as the Bank of Canada continues to target 2% inflation, and is expected to target 2% inflation in future, there is no difference between federal and provincial (or municipal) governments. Or any other corporation that issues bonds. Or any individual who borrows.
Any economist who makes a case for federal fiscal policy determining the price level must be arguing that federal fiscal policy matters because the federal government has the power to tell the Bank of Canada to stop targeting 2% inflation, and do something else instead, and that the probability it will exercise this power is non-zero, and that the probability it will exercise this power depends on federal fiscal policy.
OK. That's not implausible. If the federal debt got too high, and the federal government was unwilling to cut spending or increase taxes, so that default became a real possibility, it might order the Bank of Canada to print more money and buy federal bonds and make inflation rise above the 2% target to reduce the real value of those federal liabilities.
That is a real possibility. Unlikely at the moment, but real.
But it is an equally real possibility that the provincial debts might get too high, and that some provincial governments might be unwilling to cut spending or increase taxes, so that default became a real possibility. And the federal government might order the Bank of Canada to print more money and buy provincial bonds and make inflation rise above the 2% target to reduce the real value of those provincial liabilities.
That too is a real possibility. Unlikely at the moment, but real. And I think it is less unlikely than the federal equivalent. Some provincial governments have higher deficit and debt ratios than the federal government. [See Frances' post.] Plus, it is easier to migrate between provinces to avoid high taxes or bad government services than it is to migrate between countries. So provinces would find it harder than the federal government to raise taxes or cut spending without losing population and reducing provincial GDP. The Tiebout model bites harder at the provincial level than at the federal level.
Or it could be municipalities facing risk of default. Or corporations. Or individuals.
If you put "B(t)/P(t)" on the left hand side in the Fiscal Theory of the Price Level, what does B(t) refer to? Why should we restrict it to federal bonds? Why not add provincial bonds? Why not add municipal bonds, corporate bonds, mortgages, car loans, credit card debt, and throw everything in there? What about banks?
Ultimately, this is a political economy question. Under what circumstances would the federal government order the Bank of Canada to stop targeting 2% inflation to bail out some entity? And what entity? And what do people believe about that political economy question? Because expectations matter.
If FTPLers were serious, they would be doing political economy. And thinking about fiscal federalism. What about the Eurozone, where "national" governments are really provincial?
[OK, I promise. Last post on FTPL, for a while. But this theory needs destroying. And there actually is some Canadian content in this post, sort of.]
Worthwhile Canadian Initiative 
Decisions of the little fish are endogenous. The federal govt is large and makes unitary decisions. Therefore it does not respond to market indicators ala the wisdom of the crowds.
It’s starting to dawn on me that what you call the FTPL is completely different from what I call the FTPL (mpra.ub.uni-muenchen.de/32502). I see the FTPL as an overly narrow version of the backing theory, where the focus is on future tax streams, rather than on the subset of government assets that actually back the government’s money. So…
“are we talking about federal fiscal policy only,”
No really. Let’s say that federally issued money amounts to only 1% of government’s assets. In that case money’s value is only affected by the 1% of government assets that are earmarked as backing for money.
“or are we talking about provincial fiscal policy too?”
No, unless the central bank holds provincial bonds as backing for its money
“What about municipal fiscal policy? And why stop at governments? Lots of corporations issue bonds. And individuals borrow too.”
No, unless the central bank holds municipal, corporate, or individual bonds as backing for its money
If inflation continues at 2% forever, the price level will be infinite and money will have no purchasing power.
As suggested above, there is also the chance that irresponsible politicians will take over and speed up the process. As in Zimbabwe the money can loose all value in a few years.
Why then, so people hold money?
Of course, it is because it provides services.
Most assets depreciate and all of them might lose all value. People hold assets because they provide a return.
Now, if the Zimbabwe scenario becomes more likely, this should reduce the value of money today. It will falll in value until the services on the margin are worth enough to compensate for the greater (or sooner) risk of inflationary default.
If the quantity of money were fixed, then changes in expectations about future inflationary default would cause changes in the price level.
If we assume that a government with a bigger deficit right now slightly increases the chance of default, then it is inflationary. A lower deficit makes less likely or pushes off into the future this infllationary default. The price level falls.
The key problem with this story is that the quantity of money can change at a given price level and changes the marginal value of the services.
Bigger deficit, slight increase in chance of inflationary default, lower quantity of money, higher implicit yield, compensating for the higher chance of inflationary default.
At any given interest rate target, higher expectation of inflationary default results in more spending on inflation hedges, more nominal demand for credit, etc. The central bank tightens to avoid higher inflation–the interest rate target is higher as are all interest rates. The demand for currency falls. It looks like the interest elasticity of the demand for currency.
But actual inflation stays on target.
Maybe we should call it the fiscal theory of the nominal interest rate?
Jon: you might have a point there. Small players, like small countries, take some things as given that large players do not.
My view: the Bank of Canada is a small player. But it is alpha. All other banks, and issuers of nominal liabilities, are beta, because they promise to redeem their liabilities for a fixed amount of Bank of Canada liabilities. That makes the Bank of Canada act like a big player. It is like the “price leader” in oligopoly theory.
Mike: your version of the backing theory can provide an internally consistent answer to my question. Agreed. If the BoC only held Prince Edward Island provincial bonds as assets, then PEI would be the only province’s fiscal policy that mattered. Consistent. But WRONG! 😉 (Because the BoC earns profits, and gives them away to charity, not to its noteholders. If it wanted to, it could return more (or less) of those profits to its noteholders.
Bill: “If inflation continues at 2% forever, the price level will be infinite and money will have no purchasing power.”
You threw me with that sentence! Until I figured out what you meant. A $1 note today will have (almost) no purchasing power in the very distant future.
But I agree with what you say (unsurprisingly!).
Actually Bill, “Maybe we should call it the fiscal theory of the nominal interest rate?” that is really rather insightful. Yes!
Nick:”If the Bank of Canada gave its profits to the provinces instead, the provinces could cut tax rates a little”. Equalization payments come from general revenue so we may say that seignoriage pays for equalization. We could specifically dedicate the seignoriage to pay equalization. It would please creditists if we called it “social dividend.” It may have some media-political (mediamacro?) implications but would not change anything in the real universe. But we’re already doing it. Mere musings from a humble IO guy…
Jacques Rene: good point. If the federal government is already making transfers to the provinces, it really makes no difference whether feds or provinces get the seigniorage. We ca’t even tell the difference. Money is fungible.
It seems to me that FTPL is an attempt to build a model that has a reference point. The reference point is that government will not forever borrow money .
With an assumption that government will coast down to zero debt, all of the money represented by current debt would (eventually) move into the economy to build GDP. The amount of GDP that could be built with current debt is exactly Total Debt / tax rate. You can derive this yourself by asking “If the government was going to collect the total debt by current taxation, how much GDP would the total debt need to generate before the debt was gone.”
There are other reference points used to build models including one model that expands the debt at a rate referenced to NGDP.
Who says governments pay off debt? The UK, which basically invented modern government debt, still has bonds outstanding (they’re called Consols) from the 1700’s. They carry an interest rate of 3.5%, and another, similar bond, War Loan, carries a rate of 5%. These are true perpetual bonds; no principal is ever due unless the UK Treasury decides to redeem them. It was announced only last month that this would happen, for the entire undated bond portfolio due to continued historic low interest rates. These are bonds that have been outstanding for 250 years in the case of Consols, and 70 years in the case of War Loan. Pay off??? Says who?
Roger: it is more correct to say that the government does not run a Ponzi scheme and let the debt/GDP ratio grow without bound.
Determinant: Consols are no problem. The EPV (primary surpluses) is the present value from now until infinity.
Things only start getting hairy if the debt grows faster than the rate of interest forever. Which can happen, in principle. But FTPL doesn’t want to look at that case, so I set it aside here.
Nick, two things:
[1] Your exasperation about whether to include provincial, municipal, private debt, etc. … there is an analogous “problem” with the pure monetary theory of the price level (base money, m1, m2, etc.)
[2] The proper way to think of these matters is to begin with the understanding that monetary and fiscal policy are inextricably linked via the consolidated government budget constraint.
David. Yep.
1] Funny you mention that. I was thinking the same thing myself, after writing this post. Are us monetarists guilty of the same sin? And I think the answer is “base money”, because base money is alpha money. All the other monies are beta monies. The issuers of beta monies fix the exchange rates of beta monies to the alpha money, not vice versa. Like fixed exchange rates. If the Fed fixes the US dollar to the Canadian dollar, then the BoC decides monetary policy for the US too. BMO and TD fix their exchange rates of their dollars to the BoC dollar. Not vice versa. Alphas lead; betas follow. (Got this from Pierre Duguay, years ago. Just expanded it, and added the “alpha-beta” terminology.)
2] But that link is peanuts. It’s about 0.25% of GDP. It’s a rounding error, as government revenues go. Would it really make much difference if the BoC gave all its profits to the United Way? Or the provinces?
Nick Rowe,
So do you end up agreeing now with Scott Sumner that the unit of account is the really important kind of money?
WP: MoA is important for the price level. MoE is important for recessions. Dunno.
In the long run, price level is irrelevant so MoA is of no real import.
MoE is important for recessions. That’s important.
As the saying goes “That’s not a knife. THAT’s a knife!”
https://www.youtube.com/results?search_query=that%27s+a+knife+-+crocodile+dundee
Jacques Rene: I sort of agree. But even though the long run price level doesn’t matter at all, the long run inflation rate matters, a bit.
FTPL does not explain why currency would circulate and act as money.
If the price of money was determined by expected future surpluses, then people might be willing to hold currency as an investment of sorts, as they do today with bonds. But they don’t hold money as an investment; instead money circulates for goods and services.
We have financial instruments that promise to pay a real, price-level-independent value in the future, but real return bonds trade at a discount against cash.
It seems that the liquidity provided by cash, including the network effect of everyone else around you accepting it in exchange for goods, is itself a service provided. In a FTPL framework, this service should be accounted for as part of the surplus, even though it doesn’t show up on government books.
What does FTPL have to say about the value of bitcoins?
Majro: “What does FTPL have to say about the value of bitcoins?”
Good question. I think the answer is: “It will be zero.”!
In the US the FRB is involved now in what your text says: “the federal government was unwilling to cut spending or increase taxes, . . . it might order the Bank of Canada to print more money and buy federal bonds” – I redacted a bit here and left out the polemics of the lead in reference to debt being too high (huh?) and about the central bank’s ability to cause general-price-increase-results at this level of money-printing (in my opinion these are polemical, though they are part of the theoretic rationalization you are seeking). The FRB’s behavior has not caused the basket of prices to move much, if at all. Not known when the Zimbabwe effect starts, and don’t want the US to challenge this point in the real world.
But you have made a great point – perhaps the FRB is purchasing the wrong bonds – they should be purchasing State/Local general obligation bonds and bonds used to underwrite true asset-building projects like bridges, roadways, schools, community health clinics, etc.).
Great idea. Thanks. S/L governments spend for capital investments and spend in a Keynesian way. Much of what the federal government does is not getting into the hands of people with the greater propensity to spend (borrow-and-printed money is spent with warfighting companies, on interest payments and for many retirement checks going to people who bank them rather than add to their spending, and to cover their level-set payrolls – while health care spending actually signifies that the person of ill-health is less likely to be able to add to their spending – so these are not Keynesian outlays). The FRB should be purchasing state and local bonds!!
JF: “I redacted a bit here and left out the polemics of the lead in reference to debt being too high (huh?)..”
That was not polemical. It was a hypothetical. The Canadian federal debt is currently not too high to make default a danger, and the federal debt/GDP ratio will (very probably) be falling in future. But we can imagine a case where it did get too high, so that default vs higher inflation did become a real risk.
Would it make much difference if the Bank of Canada held provincial bonds rather than federal bonds? Not much. Slightly lower interest rates on provincial bonds, and slightly higher interest rates on federal bonds. A bit more interprovincial squabbling over which provincial bonds, and how much of each. That’s about it.
Latest news: Ontario to receive $1.25-billion increase in federal transfers
That’s around half the size of typical seigniorage profits, just for one province. Any effect on the price level?
Thanks for taking the comment seriously. The sentence you had, worked without also stating that the antecedent condition was that “debt was too high” – and I put a question there about why this is needed and more because I don’t think we know when that condition is met (a la the Rinehart-Rogoff 90% debt trigger dispute) or why it meant much to the idea of the current fiscal position to be one where it is decided to level-spend compared to the prior period but use what they would call borrowing to account for/ledger the financing reports.
Let me be polemical: you can call it borrowing at that time, if you want, but one side of the set of transaction is in fact just money ‘sent’ back to Treasury upon request. Note that since the govt’s actual activity could be seen as having been done already in a prior period, it should be obvious why its relationship to the quantity theory of money is to be questioned (which you are doing) – especially the case if the money moved over to Treasury was accounted for as financing public debt payments or even the roll-over of a debt from real spending done a decade or more before. So maybe the hypothetical is to talk about a govt that has no outstanding debt.
So if you meant the hypothetical to say: ‘if debt was already too high and the govt then had the central bank print money and send it to the Treasury simultaneously entering something on to their balance sheet’ – that is one thing. I took your hypothetical as just starting with the idea of printing so I saw the ‘debt too high’ note as a political statement intended to imply something about something (one could take it to mean a signal from you that debt is bad at some level, so take a look at this idea of avoiding it by just printing).
Governments can print money, we’ve all said that. They do not need to enter things called “debt” on to ledgers among themselves.
I’d say it was a whole lot more honest to “print” than to cut a person’s tax bill and then give them a financial asset in trade for the cash because you want to keep spending up, or increase it, or give away tax-expenditures to people, for that matte,r but don’t have the will to establish offsetting taxation. The tax-cut-and borrow approach is a wealth transfer scheme, that is for the most part all it is, and it should be avoided except in major downturns. Yes, polemics!
“The tax-cut-and borrow approach is a wealth transfer scheme, that is for the most part all it is, and it should be avoided except in major downturns.”
Nick would go further and say not even then. (Unless, I suppose, if it’s a supply-side recession that could clearly be ameliorated by suitably targeted tax cuts. Not sure what that scenario would even look like, but it’s reasonably stipulatable.)
Congratulations on the prescient article. It appears that Russia’s currency has been greatly devalued, in part over concerns of central bank financing of certain corporate debt.