« Why share-of-GDP is the correct measure for the size of government | Main | Canadian long-term interest rates are increasing as well »


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

"What is the optimal debt/GDP ratio?"

0% - Governments are like consumers, they generally don't borrow to invest, so they should only borrow when faced with an important, extraordinary non-recurring expense (medical bill, first house for people, war, economic crisis for governments).

The reverse would also apply in that if government had an extraordinary, non-recurring income stream (oil wealth, for example) running a surplus would be advisable.

But why should the optimal long-run debt/income ratio for a household be 0%?

Suppose you have a 30% debt income ratio now. Why shouldn't you just keep it at 30%? Sure, if you had lower consumption now, you could have higher consumption in the future. But why would the benefits of higher future consumption outweigh the costs of lower current consumption? Diminishing marginal utility says it's not good to vary consumption.

OK, positive real interest rates give you an incentive to pay down debt, but they also give you an incentive to keep on saving, forever (if you lived forever). Why stop at 0%? Why not keep saving?

It's a fair point, I didn't really give a reason why the optimal debt/GDP ratio shouldn't be negative. Might have to think about that one a bit.

Whatever's manageable. Whatever's necessary.

Doesn't make sense to have some optimal level of money and then target for zero debt. Not unless money growth funds all G, which seems extreme, and probably means money growth is too high. Money is special but not special enough to have zero debt. There is a demand for government debt just as there is a demand for money.

There should be a portfolio target mix for money, non-money debt, and taxes. All should be positive to fund G.

Using debt to finance private sector assets is a bad idea structurally. The target should be constrained by an ideal level of G and the mix for funding that.

The debt/GDP ratio should be kept low; governments should be allowed to run cyclical deficits and surpluses within some range.

No idea what the ideal number is though. 30 per cent feels good somehow.


There is a demand for money (people want to hold it as a medium of exchange even when the real rate of interest on money is below other assets, or even very negative because of inflation). So the government should produce money. It can earn a profit from producing money (seigniorage, or the inflation tax).

If there's a demand for government bonds, so that people want to hold some government bonds even when the interest rate on bonds is very low, you can make a similar argument that the government should produce bonds. And it can earn a profit by producing bonds that pay a rate of interest below the market, if government bonds, like money, are "special" in some way, like money. Like banks, if the government can borrow at 3% (issuing bonds), and lend at 5% (invest the proceeds in private financial assets) it can make a 2% profit (ignoring costs).

But the government could produce money and bonds while having 0 NET debt. Or negative net debt, or positive. Just depends on whether it buys any offsetting financial assets, and how much.

The bigger the net debt, the higher taxes will need to be, to pay for government spending plus the interest on the net debt. But we need temporarily high taxes to reduce the net debt and get lower taxes in future.

So, what's the optimal NET debt/GDP ratio? Dunno.

My general rule of thumb is this -- if you're borrowing to spend on something from which you intend to generate enough revenue to pay the principal and interest (and ideally some profit) you have a case to be taken seriously. This applies more easily to the private sector than government however because the benefits we derive from government spending are difficult to measure. But the invisible losses we suffer from taxation and crowding out are equally difficult to measure, so arguably they cancel...? Either way, I prefer government debt to be 0 because I have no control over it. I don't need to hold an election and argue against inbred ideology to pay off a credit card, should I have chosen to run that debt to begin with. Another reason is more political than economic, I am one of those people who mistrust people in power. Even if you could conceivably create a responsible and reasonable powerful central government, I wouldn't take the risk.

By the way anon, I should have said that I'm mostly agreeing with you. Except that your argument, though very reasonable, points to an optimal GROSS debt/GDP ratio, not an optimal NET debt/GDP ratio.

Isn't asking what should the debt/GDP ratio be looking at the problem backwards? Wouldn't it be better to ask what deficit/GDP ratio won't cause significant "crowding out" of private investment AND can be usefully invested in public goods? Since this ratio will vary over the business cycle, the debt/GDP ratio would follow.

But since that is probably too complex of a problem for economists to solve or citizens to understand, another way of addressing the problem is that the operational budget should be balanced over the business cycle and that the capital budget should be limited to what can be serviced given a country's preferences about taxation and other types of spending.

And since that is probably too complex still, the answer is 25% of GDP. Just because.

0% - Governments are like consumers, they generally don't borrow to invest

Nonsense. Governments do borrow to invest all the time - just look at the US muni market (excluding Gullyvornia).

Nick mentioned in his previous post that there is a market for government debt.

Government debt serves as a baseline for expected 'risk free' returns as well as providing a tool for monetary policy.

About two years ago I remember that the supply of government debt was becoming so thin and demand so robust that there were investigations into market manipulation.

It's important to the consider the role government debt plays in greasing credit markets. Any argument that debt enables irresponsible government is besides the point and more political than economic. I consider the last 30 years of governing and government debt to be very responsible, especially compared to our international peers.

In terms of no 'net debt' I really don't know how comfortable I feel where the government owns private assets. Say the government debt is $400 billion, where would it park it? Stocks? Bonds of other countries? This investment strategy hasn't worked so well in other countries (China, oil SWFs), I'd rather the government park it's assets with Canadians in the form of lower taxes or higher spending and positive net debt.

The reason to have interest paying debt as well as money is that there is a natural demand for low ('zero') risk financial assets for durations longer than the duration of money (which effectively is zero).

I've suggested positive net debt, including money and interest paying debt.

(It seems inefficient for the government to be involved as a permanent intermediary between its own liabilities and private sector assets - this is potentially profitable but a questionable use of the government's risk free 'brand')

When the real rate of interest on money is negative, money issued by the central bank becomes an economic asset rather than a liability. So there is an offset between interest paying debt and money in that sense. It would seem to argue in favor of issuing interest bearing debt at least up to the point of offsetting ('using') the benefit of issued money as an asset.

I'm lost, could someone explain why surpluses lead to communism? Couldn't they simply lend any surplus out to other nations in order to bank their money for times like this when they need it?

anon, that's not a reasonable answer in my opinion. The banks also want to lend with minimal risk, that doesn't mean everyone with good credit should feel compelled to borrow for their sake. Lots of people want lots of things, if investors want low risk, let them create their own low-risk investment instead of wasting it on the government and expropriating my labour to make them whole.


implicit in my answer is that some level of debt is targeted that replaces what otherwise would be taxes; i.e. there is an appetite for low risk debt that is a legitimate substitute for taxes; the level of G is a separate problem; G and money and debt and taxes must all be kept under control - portfolio size (G) and funding mix (money, debt, taxes)

Traciatim, in the limit if surpluses are large enough then eventually the government ends-up owning everything by collecting more in taxes than it spends on running the gov't, providing public goods and services, and servicing debt.

I might be wrong on this, but I think that it is possible to run surpluses forever without seizing the command heights so long as the average tax rate is less than the rate of private sector growth.

anon, taxes don't get repaid. How can these be comparable sources of revenue?

oops ... commanding heights ...

Traciatim: if other countries' governments weren't fiscally conservative, and had debts, the government of Canada could lend to those other countries. But if all countries tried to do the same thing, it couldn't work.

Mark: You presumably meant to say "Suppose the government debt is MINUS $400 billion"?

I think that if we accept Mark's argument, that we don't want the government to be owning private assets (except perhaps in special cases), then net debt and gross debt come to the same thing. And we could then use anon's argument about a demand for government debt to get, in principle, an optimal debt/GDP ratio.

pointbite @ 1.05: But we might accept there are some things the government can do better than the private sector. That's why we have a government. And issuing debt might be one of them. The underlying theory might be that the government has a comparative advantage in some sort of ultimate risk-pooling, both across people at a given time, and across generations, because it has the power to tax. We are all, as taxpayers, like Lloyds "names", with unlimited liability.

But this is perhaps the very point about unlimited liability that causes pointbite and anon of the moment the most concern. Even if there is no optimal debt/GDP ratio, we need to draw some line in the sand, even if we recognise it's arbitrary, to put some constraint on government. So we look for some Schelling focal point -- a "magic number" like 0, or 25% of GDP, if we have become accustomed to thinking that 25% is the target. Sure it's arbitrary, but then so is the 49th parallel arbitrary as the US/Canada border. But even if we do need some line in the sand, I would like to feel that we had drawn it in at least roughly the right place, even if the exact point is arbitrary. 100% is a nice round number too? (Except it's only a round number if we measure time in years, since we are comparing a stock of debt with a flow of GDP.)

Anon of the moment:
"Isn't asking what should the debt/GDP ratio be looking at the problem backwards? Wouldn't it be better to ask what deficit/GDP ratio won't cause significant "crowding out" of private investment AND can be usefully invested in public goods? Since this ratio will vary over the business cycle, the debt/GDP ratio would follow."

If we are asking about the *long run* debt GDP/ratio, budget deficits always cause 100% crowding out in the long run (either of C+I in a closed economy, or Net eXports in an open).

In an economy where Ricardian Equivalence is true, the level of government debt/GDP ratio won't affect the long run capital stock and national wealth. In an overlapping generations model, where Ricardian Equivalence is false, because future generations will bear some of the tax burden, a higher debt/GDP ratio will generally cause a lower capital stock. But sometimes that's a good thing. In particular, if the interest rate on the debt is lower than the long run growth rate of GDP (measure both in real or both in nominal terms), the government can run a permanent stable Ponzi scheme on the national debt (borrow to pay the interest, and the debt still grows more slowly than GDP), and make every generation better off.

Patrick, that theory stretches reality to the point of near absurdity, again in my opinion. If we were smart enough to stop spending, it's not unreasonable to assume we would be smart enough to stop taxing when the debt is gone. And regardless that theory assumes government spends money in ways that increase their ownership of assets, which is not usually the case. The government doesn't accumulate $200 billion worth of assets every year. Or perhaps the massive government intervention in the economy that would result from an explosion of government expenditures would cripple the competitiveness of the economy and the tax revenues along with it. Either way, deciding on the least damaging way to spend too much money is a much better problem than balancing bankruptcy on the tip of a pin. It reminds me of a hockey team debating which of their all-star goalies to put in nets...

pointbite, government issued debt is repaid with future taxes.

I think the point is that there is a demand for risk free investment, and government debt provides that (well, first world government debt sorta does), so why shouldn't the government use that do the equivalent of optimizing it's capital structure?


"But we might accept there are some things the government can do better than the private sector. That's why we have a government. And issuing debt might be one of them. "

I can accept the premise that government can do some things better than the private sector, and we may actually want those things, but I don't agree that debt is one of them. The idea that someone or some institution can be "better" at borrowing money doesn't sit well with me, I don't even know what that means. A maid in a rich person's house has a lower risk of default because she can always steal from her employer, but that doesn't mean she's doing him any favours. What matters to me is not the maid's risk of default or the loan shark's income... it's where the money ultimately comes from, where it's spent and what else could the economy have done if that money were it left in more responsible hands (even if against their will).


given the assumption of an ideal non-zero debt/GDP ratio, and trend growth in GDP, the trend level of debt never gets repaid - the same as for the trend level of money

that's why you don't want 100 per cent taxes funding G

pointbite: Yes, it is a weird concept, and I'm trying to get my head around it too.

Here's an example. Suppose there are 365 of us, and our birthdays are spread out over the year. Each of us wants to save money for a big birthday present to ourselves. (Sorry, I can't think of a better story). And we want our savings to be as safe as possible.

One way to make our savings as safe as possible would be for all 365 of us to join a savings club and agree to be jointly and severally liable (is that the right legal jargon?) for the liabilities of the club. But it might be hard for us as individuals to enforce the rules of the club. We can't incarcerate people for non-payment of club dues.

You can think of government debt as like the liabilities of such a club, except that the government manages the club, and enforces the rules. And joining the club is compulsory.

Open to abuse? Yes, absolutely. But could in principle do something we couldn't otherwise do, but might want to do? Yes.

anon, the fact money is debt today doesn't mean it has to be that way.


I don't have time for your cartoon.

Do you have a point?

(other than your statement above, which is incoherent)

Sims thinks government debt is like equity

So it should be permanent

Isn't the technical term for a compulsory club a "prison"?

the fact that debt equates to the PV of future surpluses is only the expression of a funding option

the government can choose to issue more debt instead

pointbite: I was watching that video you linked. It was doing OK till about halfway through, then it totally lost it, in the P/(P+I) bit, so I stopped watching.

Do you know who produced it?

Nick, I think there are credits at the end, I watched it a long time ago. I would like to read your criticisms of the clip if you have any worth discussing, it's very popular so I'm sure it could generate lots of interest in this blog.

anon, the point of that cartoon is that debt can't be repaid because all the money used to repay it is also debt, if money circulated permanently that wouldn't be the case.

Pointbite, you say "the fact money is debt today doesn't mean it has to be that way".

A couple things, first of all if the following two statemtenst are true:

1) the government collects taxes;
2) the government can create as much money as it wants;

then yes, it just is that way.

Secondly, going back to a system in which the second of those two statements is false is a horrible idea.

I'll try and watch the video later

But repayment of debt only requires a redistribution of "money debt" from taxpayers to debt holders

The fact that money is debt is no constraint on this

And total debt (money plus non money) does shrink

pointbite: the weakest point in the video at about the 20 minute mark, is where it said the money supply has to increase at the rate of interest so people can pay the interest on the debt. That's when I stopped watching.

This is wrong, because it confuses a stock (the stock of money, and the principal on the debt) with a flow (the interest payments). In principle, a single $1 could pay all the interest owed each year, if it circulated at a high enough velocity.

A minor weakness was its saying that we all owe banks money. True, but banks all owe us money. If we include the banks' capital, which is what banks owe to their shareholders, it's a wash.

There's also a hint of the idea that banks are the reason behind interest, and that by creating money banks earn an illegitimate profit. Interest would exist even without banks, and even without money. And if banks have a monopoly, they can earn monopoly profits by monetising assets. But then if you have a monopoly on any good, you can maybe earn monopoly profits. If banks are competitive, competition will bid up the interest rate on deposits. (The film gave the impression that interest on deposits was some sort of social compact between banks and citizens to split the loot between them.)

I like to think of banks as converting illiquid financial assets into liquid media of exchange. That's generally a good thing. In my idealised perfect world, all my assets, real and financial, would be perfectly liquid, and usable as media of exchange.

On the money=debt theme: most money is debt (I would exclude base money, but that's another story); but most debt is not money. Total debt is much larger than total money.

But the idea that the quantity of gross debt might be linked to the supply of money is interesting nevertheless. We can imagine a world in which governments had zero debt, all businesses were run by owners who used their own capital, and all households had the same desired savings, so their were no consumer loans. So we can imagine a world where nobody would want to borrow and lend. But there would still be a demand for money as a medium of exchange, because people are still buying and selling goods. There is intra-temporal trade, and no inter-temporal trade (borrowing, lending, gross debt).

My guess, without watching the whole film, is that it's an offshoot of the Social Credit movement. Which, amazingly, still has an underground presence in Canada. (Just glad to see the bankers did not look like stereotypical Jews, which they used to in Creditiste flyers I remember from the early 1980s).

On balance though, there's something to be said for reading those disparaged as "monetary cranks". At worst, it forces you to think through your model of money. At best, there may be a grain or two of new insights in what they say.

Back on clubs vs prisons: sure, a compulsory club can be seen as a prison. But I don't see any difference between compulsory clubs to pay for public goods like police etc., and compulsory lending clubs.

I saw the first 30 minutes; its all I could stand.

It's a childish piece of Austrian propaganda.

Reminds me of 1950's style bomb shelter films - horribly bland and droning narration aimed at hypnotizing the listener into submission.

Agree with comment above re interest payments as flows. Not surprising the film got that wrong; its the only conceptually sophisticated point attempted.

A couple of points on fiat currency are nearly correct, but the thing as a whole belongs in the ash can.

Hey Nick, watched the call-in show (Goldhawk Live) Sunday ... must say that the crazies were out in full force, and some in-the-tank Liberal voters (or perhaps even part of the apparatus). You did well ... what’s your accent, if you don’t mind me asking? I’d guess Yorkshire or Lancashire (though mentioning those two side-by-side can be fighting words in some parts).

For me, I think the government has handled the situation as best as you could expect any government composed of mortal and fallable beings to have handled it. Sure, things could’ve been done that would’ve been better here or there, but on the whole, they’ve done a decent job. I would be horrified if I were an American right now, watching their government rack up a deficit that’s 12% of GDP, on a track to a debt load over 100% of GDP. That should scare them, and us.


Yes, those are all valid points. The film is definitely simplistic, but it's an introduction to a topic that most people have never thought about before, so I can forgive them for it.

Regarding clubs vs prisons, I find it a little disheartening the difference between a prison and our government is so insignificant that this discussion is even necessary, but nevertheless... I see an enormous difference, especially when the context is so vague.

It's not the existence of the lending club that bothers me, is the compulsory nature of it. I think the distinction is really the difference between a democracy and a constitutional republic. If you support a democracy then you support a tyranny of the majority -- if you can convince the majority of my neighbors to steal my car, I have no right to protest -- there aren't really any property rights. If you support a republic certain things like police and courts are necessary evils (the existence of which are not very controversial anyway) and beyond some un-ambiguous, legally defined, narrow and generally agreed upon boundaries the "club" has no power of compulsion whatsoever. The rules have limits. We can't just decide one day to tax AIG executives at 90% after some temper tantrum. We shouldn't be protecting the club from irresponsible people, we should be protecting the people from irresponsible clubs. Not every club is bad, I will admit that so don't use that argument, but that determination should be mine to make, not some committee in a dark room that doesn't give a damn about my interests.

The difference is the difference between freedom and slavery. Nobody should have the right to create a club, on a whim, and compel me to join it. If you define tyranny as exercising a power that nobody should be allowed to posses, that Nick, is tyranical.

Here's the source of it:


anon @6.08: Thanks for the link to the producer.

anon @ 5.42: same "anon"? How about a nom de plume?

It's not Austrian; it's Creditiste. From reading the producer's link, he holds a monetary theory of the rate of interest (Austrians hold a real theory), and wants interest free loans from the Bank of Canada, which is a signature tune of Major Douglas and Social Credit. Of course, the Bank of Canada charges interest to the government of Canada, and earns a profit, but repays all that profit to the government, so it's a wash. And if the banking system is competitive there aren't any monopoly profits from bank-money anyway.

KRB: Thanks! But I wished I could have switched places with Stephen, so he could have handled the EI issue better than I did. It was the other big topic.

I'm always surprised at my accent when I hear myself recorded. My God! Do I really talk like that? It's a bastardised mix of Herfordshire/Bedfordshire border rural remnants plus London overspill plus public school plus 3 years in Scotland plus 1 year US and Australia plus 30 years in Canada. With influences diminishing as I age and get set in my ways. But Yorks/Lancs only by convex combinations of the above!

pointbite: that wasn't what I was trying to say. What I meant was there is a big difference between a voluntary club and a compulsory one, but I can't see much difference between a compulsory club for borrowing and lending and a compulsory club for the provision of other goods.


What I was trying to explain is that there is a difference, but it's qualitative. Even the most idealist libertarian supports courts and police, those "public goods" or whatever you want to call them are not in dispute, but once you start talking about compulsory lending you have to ask how that money is being provisioned and spent. It's reasonable for me to accept a police officer enforcing a reasonable law, but it's not reasonable for me to accept my neighbor can steal my car (or other property, like income) because he held a vote in his yard. Once is consistent with freedom, the other is not. It's not reasonable for someone to accept a compulsory lending club in principle without more context. Hence the qualitative differences between the type of compulsory things you find in a republic and a democracy. Not all compulsory clubs are created equal.

Summary -- you can't extrapolate from "we have compulsory courts" to "we have compulsory graduated income tax to fund a bailout of GM". Both compulsory, but enormous differences between them.

What stimulus? Unless you count the loans to bankrupt US car companies, federal government spending this year has not even kept up with inflation.

pointbite: I think you'll be sorely disappointed to discover that any society is a 'prison' or a tyranny in the manner you describe. The only way to avoid it would be to live in total isolation from other individuals, which is not very practical.

This is wrong, because it confuses a stock (the stock of money, and the principal on the debt) with a flow (the interest payments). In principle, a single $1 could pay all the interest owed each year, if it circulated at a high enough velocity.

This hints a very important rub: if investments are profitable, then the increases in velocity will cover the interest payments (and more) and the money supply need not expand. What about when ROI < i?

Andrew, varying degrees of tyranny, yes.

Jon: the velocity of circulation of money is the number of times per year that a dollar (the medium of exchange) changes hands. That's quite different from the profitability of investment.

Nicks point about velocity is very interesting. Maybe all those posts about demand for money are just sinking in...

So is it fair to say that an increase in the demand for money slows the velocity of circulation and can eventually cause a recession? And would it be fair to say that a bursting credit and asset bubble is pretty well guaranteed to slow the velocity of circulation significantly as everyone tries to turn assets into cash and generally snaps up every spare dime to fill the credit hole (made worse with the debt deflation dynamic).

I'd hope by now that you'd know that I know what velocity is... so you missed the point entirely.

As per Patrick, I'd like to hear more about the relationship between the demand for money and the velocity of money.

see this:


Sorry Jon: but I'm still not getting the point. If the rate of return on an investment exceeds the rate of interest, then the investment will pay for the loan (eventually). If not, it won't. But I don't see the relation between that and the velocity of money.

anon and Patrick: if you define velocity as "income velocity", as in MV=PY, then a (desired) fall in velocity is the same as an increase in the demand for money, for a given price level P and real income Y. Two different ways of saying the same thing.

Yes, a fall in desired velocity will cause a fall in aggregate demand (PY) and will cause a recession (given sticky prices). But that leaves open the question of why desired velocity fell. Did it just happen, or was it caused by something else, like a fall in the rate of interest?

anon @5.02 Yes, we are in a much better fiscal situation than the UK. But, IIRC, the Brits have had a debt/GDP ratio of 200% 3 times in the past: after WW2; after WW1; and after the Napoleonic wars.

Bernanke today:

"In particular, over the longer term, achieving fiscal sustainability--defined, for example, as a situation in which the ratios of government debt and interest payments to GDP are stable or declining, and tax rates are not so high as to impede economic growth--requires that spending and budget deficits be well controlled."


If our government is providing all the necessary services (i.e. health, education, infrastructure, social services and policing) with a satisfactory quality I don't see why we cannot have surpluses forever. Any accumulated surpluses could be invested in foreign countries bonds earning us taxpayers a nice income and work as an emergency fund for rainy days. If the accumulated surplus becomes too big (say > 20% GDP) then taxes could be lowered, but I don't foresee that day coming any time soon :).

The comments to this entry are closed.

Search this site

  • Google

Blog powered by Typepad