It's a common theme that the financial crisis was caused by too much debt, relative to our incomes. So if we want to get out of the crisis, we need to reduce our debt/income ratios. How?
For an individual, the answer is easy and obvious: spend less, save more, and pay down debt. But for a country as a whole this won't work. If everybody tries to save more, and spend less, income falls (because one person's spending is another person's income). The paradox of thrift tells us that if investment, net exports, and the government deficit stay the same, savings stays the same, so that the attempt to save more fails. The paradox of debt is like the paradox of thrift on steroids. If we each attempt to reduce our debt/income ratios by saving more, income falls, and so our debt/income ratios rise, making the problem worse, not better.
It's different in an open economy, because if we save more and spend less we may import less, and so pay off debt to foreigners. In a very small open economy, where nearly all changes in spending are changes in imports, it's just like the individual, where our own saving doesn't affect our current income. But if the rest of the world is facing the same problem, the paradox of debt just reappears at the world level.
If we could all save more, consume less, and invest more, so that aggregate demand stayed the same, current income would stay the same too (and future income would eventually rise as the stock of capital allowed us to produce more). But most investment is done by firms, and firms finance investment by borrowing, and most of that borrowing takes the form of debt (not shares). So household debt might fall, but firms' debt (plus shares) would rise, by an equal amount.
It's just the same if increased household saving is replaced by increased government deficits. Current income stays the same, but household debt is replaced with government debt, so the total debt stays the same.
There are four ways to reduce the average debt/income ratio in a country.
1. Households who are net debtors save more; households who are net creditors save less. "We owe it to ourselves" is true; "some of us owe it to others of us" is also true, and more useful.
2. We replace debt with equity. Instead of Mary lending Peter $100 to buy a house, she buys the house, and lets him rent it. Firms issue shares instead of debt, and people own "risky" shares instead of nice safe debt.
3. We disintermediate. Instead of households lending $100 to banks, and banks lending $100 to firms, which makes total debt $200, households lend $100 directly to firms, which makes total debt $100.
4. We default on our debts.
OK, which one are we going to do? Will it happen of its own accord? If not, how do we help or make it happen.
I don't have the answers.
Worthwhile Canadian Initiative 
I don’t have the answers.
That’s not surprising since you’re essentially trying to decide if it makes more sense to keep the drapes open or closed while the house is burning down. As long as we cling to a consumption economy it’s not possible to reduce the debt/income ratio. Every proposed solution I’ve ever seen doesn’t accomplish anything other than shifting the problem somewhere else.
But Robert: Y=C+I+G+NX. If we earn Yncome from producing output, and we don’t Consume it, we must either Invest it, have the Government buy it, or Net eXport it. So I did consider all three alternatives to a “consumption economy”, and they seemed just to shift the problem elsewhere. My four “solutions” however, did not shift the problems elsewhere, and did not require shifting away from consumption.
5) Let the long passage of time wear down debts as firms claw their way back into solvency. i.e. Japan.
6) Inflation.
Here’s a question I have about debt and the Y=C+I+G+NX identity. Say I borrow 200k from the bank and buy
a newly built 200k house. That gets added to I, right? Is the value of the debt reflected anywhere? And
what happens to Y if I have to default on the loan next year – do one of the variables get reduced?
tc: your (6) is a good addition, if we assume it’s an unexpected inflation. (An expected inflation will just raise nominal interest rates.)
I had to think about your (5). I think I would file it under my (2), since when a firm pays down debt out of retained earnings (so has the same total assets), it’s really like replacing debt with equity, since the shares are now worth more. But it does require that people are willing to hold a higher percentage of their wealth in “risky” shares. (Modigliani-Miller says it wouldn’t matter of course, and if we modify MM by introducing costly default, a higher percentage of shares relative to debt in the average portfolio actually increases safety.
Maybe we should tax debt, and subsidise equity, if a high debt/equity ratio increases aggregate risk?
Maybe when Robert above said “consumption economy” he was meaning an economy where we work, produce, and consume, and was advocating an alternative where we all work less, produce less, and consume less. Of course, in the long run, there’s no problem with doing that. If people decide they don’t need all that stuff, they stop buying it (so AS falls). But then they should also decide, if they don’t need all that stuff, they don’t need to work as hard either, so labour supply falls, and AS falls as well. So it doesn’t create a recession (except in the stupid technical sense). Or, rather, it doesn’t create unemployment.
But if you suddenly decide to live the hippy life, after going deep into debt to buy a load of stuff, other people (the ones who lent you the money, and may be relying on you to keep them in their old age), might not be too happy.
tc: sorry. I forgot about your question:
“Here’s a question I have about debt and the Y=C+I+G+NX identity. Say I borrow 200k from the bank and buy a newly built 200k house. That gets added to I, right? YES Is the value of the debt reflected anywhere? NO, NOT IN THAT IDENTITY, WHICH IS ABOUT THE PRODUCTION AND SALE OF NEWLY PRODUCED GOODS AND SERVICES And what happens to Y if I have to default on the loan next year – do one of the variables get reduced?” NOTHING HAPPENS TO Y DIRECTLY, THOUGH THERE MIGHT BE INDIRECT EFFECTS IF YOUR DEFAULT CHANGES PEOPLE’S EXPENDITURE PLANS.
Debt and default doesn’t show up on the Y=C+I+G+NX identity. You really have to look at the balance sheets (assets and liabilities) to see it.
“Households who are net debtors save more; households who are net creditors save less”
Interesting way of looking at it – financial intermediation shrinks – makes sense since a lot of the problem is in the distribution of “we owe it to ourselves”. (Damn bankers.)
“We replace debt with equity”
Also interesting, but isn’t corporate capital structure a different type of problem than household debt distribution? Where does the MM hypothesis factor into this interpretation?
“We disintermediate”
Also interesting – surely the growth of debt through multiple phases of horizontal intermediation is a different problem than the growth of debt through leveraging vertically in a particular capital structure.
Nick, thanks for the answers.
Expected inflation will of course affect the rates of loans made in the future. But what about loans made in the past? If lenders thought we had solved macro and would be in the Great Moderation forever, then any rise in inflation would be unexpected to their past selves.
If debt doesn’t show up in the identity, does it show up in GDP calculations? Just going off the official statistics, in 2008 the US showed increases in GDP and productivity, which leads guys like Casey Mulligan to say weird things (like unemployment was actually due to people who don’t want to work). Or is GDP systematically flawed, in that it makes countries that are running up debts look richer than they actually are?
JKH: regarding MM, see above at my 12.53 comment (We were probably writing at the same time). Under MM assumptions (no default) it doesn’t matter. But if we assume default, and especially if we assume costly default, then MM is violated. There’s a weird sort of paradox in debt/equity ratios. For the individual, a high debt/equity ratio in the portfolio reduces risk. In aggregate, a high debt/equity ratio increases risk (of financial crises). Still need to get my head around this.
“Also interesting – surely the growth of debt through multiple phases of horizontal intermediation is a different problem than the growth of debt through leveraging vertically in a particular capital structure.” I’m not sure I’m following you there (and/or I can’t get my head around it). As I was writing that stuff, deleveraging seemed like a combination of disintermediation and replacing debt with equity. But I wasn’t quite sure.
Off skating again. Back Sunday.
tc: yes. An unexpected change in inflation after the debt was signed, on a fixed interest loan, will reduce real debt.
If you borrow from abroad, and invest the proceeds in domestic factories, Gross DOMESTIC Product will rise (because you are producing more stuff within Canada), but Gross NATIONAL Product won’t rise, or won’t rise as much, because you are having to pay extra interest to the foreigners who lent you the money to build the factory, so part of the output of that factory is deemed to be produced by foreigners, not by Canadians.
Not much of an issue for Canada at the moment, because net foreign debt is very small. Bigger issue for other countries.
Definitely off skating now!
I expect 4. It will allow the Spender Country/ Saver Country Symbiosis to continue, and doesn’t ask anyone to change their behavior. The other three options simply ask too much of human beings.
People will eventually be forced not on their own accord to make the hard choices.
The net debtors will have to cut back when they realize mastercard, visa, and the big five banks will no longer subsidize their lifestyle. I think once the prices and interest rates reach low enough levels, the frugal net creditors will get spending and save less. How low is low enough becomes the question?
As for defaulting on debt, I don’t think the government would allow this to occur on a mass scale given the economic consequences and psychological impact. If I am not mistaken the CMHC exists with mortgage insurance for Banks to insure them against home buyers defaulting on their debt. Based on what has occurred already in Canada and the US it looks like business (esp. those in vote rich BC, S. Ontario and Quebec) will get something to cushion the fall (if not to prevent it all together in the SR). With a low debt-to-GDP ratio compared to that of the US, Italy and Japan I don’t believe the government itself will have to resort to option 4.
Because the Harper government could fall at any moment, I am going to say the government will do almost anything in its power to avoid option 6 of raising inflation. Most people (myself included with an economics background) don’t think in real terms but rather in nominal terms. The Tories will want to make sure that inflation is not high. People hate inflation and governments are scared of it. The I-word may be the only word worse than the D-word. People are unhappy with paying extra for gas prices alone, imagine if that was economy wide?? With high inflation Harper could lose the next election d there’s a good chance he’d be done as leader as well. Humans are not completely rational actors and even if the classical dichotomy was correct and the AS curve perfectly vertical, people will still complain when prices increase no matter what other economic variables do. If people don’t want it, governments will do what it takes to win votes.
A case in point is the Liberal’s Greenshift Plan. It was successfully attacked for the very reason it would increase the price of goods!
Hopefully the BoC will continue to pursue its 1-3 percent inflation band. Nick you probably know more about this, but the BoC’s agreement for the 1-3 percent inflation band isn’t up for renewal for a few years still and there is currently no reason to suspect the BoC will want to change that.
Sorry for the excessive (and tangential) rant about inflation, but it’s just something I am not a fan of. All that said, I am going to be optimistic and no one tries to pull a fast one lower the debt-to-income with an increase in inflation.
‘The paradox of debt is like the paradox of thrift on steroids.’
I like that.
Has anyone considered stagflation?
Here’s an article that may interest you:
http://www.independent.co.uk/news/business/news/obamas-team-turn-to-eu-bank-for-inspiration-1514983.html
The old mantra by businesses/economists was ‘we need to get consumers spending’. What got some people into this mess, imho.
The new mantra is ‘we need to get governments spending’.
I suppose the U.S. government (taxpayers) would be considered to be intermediating or adding value if they follow through.
Here’s a crazy Idea I’ve been mulling:
Suppose the government gave everyone $10,000 vouchers that can only be used to pay off debts or to invest in your RRSP. And suppose the banks who receive these vouchers are required to deposit them with the Bank of Canada, where they are counted as reserves. Everyone has less debt or more savings, and the banks have improved their reserve ratios, and all this has only cost the paper and ink to print the vouchers. Life is beautiful, the financial crisis is over, and everyone can go back to watching Hockey Night in Canada.
What’s wrong with this picture?
I’ve been thinking about this too.
My thought is that there needs to be an increase in C and a decrease in savings in creditor (i.e. raise the standard of living of the poor) nations and a decrease in C in debtor nations which gets redirected to savings (i.e repent our profligate ways). The increase in C in creditor nations could drive an increase in NX in debtor nations.
So, and this is going to sound crazy, the smart thing to do is to send a few trillion in stimulus to China, India, South Africa, Brazil, etc. as long as they promise to give it directly to their people to spend.
Yeah. I know. Not going to happen.
Hi Nick, I was thinking about your post with respect to the quantity theory of money. Since I’ve never really studied macroeconomics, I’ve probably missed something obvious here, which I encourage you to point out, but I just wanted to throw my thoughts out there so someone could point out the flaws.
My understanding is that the theory says that our gross income, expressed as price x quantity, is equal to the velocity of money times the quantity of money, which intuitively makes sense.
Then, we can separate the money supply into two components. The piece that doesn’t need interest paid on it (e.g. currency notes), and the part, created through fractional reserve banking, that does need interest paid on it (e.g. credit, or debt).
You could write it as Gross Income = Velocity of Money x (Currency + Debt).
If we hold the velocity of money constant for the purpose of the discussion, then the ratio of debt to income depends on the proportion of the money supply created as currency vs. the portion created as debt. You could view the ratio of these two elements as our Economy-wide leverage ratio.
So, like you say in your post, reducing the debt component of the money supply would decrease income, but unlike your post, it would still reduce the debt / income ratio. This makes sense again as you can imagine that if we paid off all our debts, then we would still have income generated by the circulation of the remaining currency. A model where debt repayment leads to a greater decline in income would lead to negative income before debts are eliminated which doesn’t make sense to me.
So, if we wanted to decrease our debt to income ratio (keeping velocity constant – all else equal a rise in velocity would decrease the debt/income ratio and a decline would increase it), we would need to either increase the currency component of the money supply (what you might call inflation) or decrease the debt (credit) component of the money supply (either through repayment, or more rapidly through default).
How high a leverage, or ratio of debt to currency the economy could support would depend on interest rates, just like the leverage a company can support depends on the interest rate is pays.
I’d guess that over a boom period, the leverage (ratio of credit to currency) would increase as banks increase their own leverage. At some point the leverage would become dangerously high, threatening an implosion. At this point central banks could raise rates to trigger this implosion in a controlled fashion, creating the defaults that would wipe out enough debt (credit) to reset the leverage ratio to a safer level and then, once enough leverage was wiped out, they could lower interest rates and start the cycle again.
Alternatively, they could postpone the implosion by sending interest rates lower and lower, since the lower interest rates go, the higher the leverage (and the higher the debt to income) that can be sustained. But this game reaches an end once interest rates start to bump up close to a zero bound and the leverage is so high that people are still defaulting. Now you get the same implosion of debt/credit that you would have had earlier, only its a lot bigger and you can’t help start a recovery by lowering rates so you need to wait a lot longer / work a lot harder to get your leverage ratio back to a reasonable level.
Thoughts?
Don: (4) (default) is what has been happening, but default is costly (see my previous post on default), so I hope we don’t go this way.
Mark: Yes, (1) (borrowers being credit-constrained and saving more; lenders responding to low interest rates and spending more) is the best way to reduce debt while avoiding a recession. To some extent this will happen, or is already happening. The danger is that lenders will see the losses on their loans, from default, or risk of default, and will try to save more to recover their losses.
IIRC, the BoC’s 2% inflation target comes up for renewal in 2010 (or 2011?) I expect them to keep to 2%, but may switch to a price level path target rather than an inflation target (i.e. they try to correct past errors to make the future price level more predictable).
Alex: Banks won’t like your idea, if they can’t do anything with the vouchers it’s effectively a tax on banks. Let’s change your idea: suppose the government borrows $1,000 per person, hands it out, and requires them to pay off loans or add it to an RSP. (It’s the same as what would happen if everybody saved a $1,000 tax cut.) It replaces private debt with government debt. It doesn’t reduce total debt, but if the government is better able to handle the debt than individuals, it might be an improvement. Future taxpayers might not be happy, but it’s not obviously a bad thing.
Dee: stagflation normally comes under two circumstances: supply shocks; and when expected inflation is increasing but the BoC tightens monetary policy to keep actual inflation lower than expected inflation. I don’t think it will happen (and I hope not).
Can you think of the government as an intermediary? Let’s try it: The government borrows from today’s lenders, and makes a loan to future tax payers. It then takes the money the future taxpayers have borrowed and spends it on stuff. The future taxpayers then have to repay the loan. No. Thinking that way doesn’t help me keep my head straight.
Patrick: If China starts spending more, and the US starts saving more, that is the equivalent to (1) on a global scale. (The US is a debtor, and China a creditor nation). It’s what one hopes will happen. But I don’t see why the US should pay China to spend more. That doesn’t help, since the US will borrow from China (so the debt problem gets worse), then let China consume the stuff.
Declan: that one hurts my brain, trying to think it through! I run into trouble at the beginning, since money (by definition) is a medium of exchange, and debt isn’t. (Debt has to be switched into money before people can spend it). I may have another go at it if and when my head clears.
Nick?
What about a demand shock? BoC has some room to maneuver but what if those are exhausted?
I think an intermediary would work. Remember war bonds? As long as the lending is done amongst the citizens. How much did the President manage to raise for his campaign through small contributions?
Interesting times.
Dee: a (positive) demand shock would be nice right now. But where is one going to come from? We may get lucky, but I wouldn’t count on it.
War bonds were a little before my time (sorry). But that sounds like a variant of an increase in government spending financed by deficit and borrowing. The only difference between war bond drives and regular bond sales is that financial intermediaries were skipped (I think). The government could do the same with Canada Savings Bonds, which are typically held by individuals, not banks (I think).
Would a sudden demand shock be good? Wages haven’t moved. People have been using debt/credit to supplement their incomes for basic staples (food and fuel). Cost of debt/credit increase along with a job loss could result in defaults.
I still see a possibility of stagflation but will concede defeat, for now.;^)
‘But that sounds like a variant of an increase in government spending financed by deficit and borrowing.’
Of course it’s a variant but I imagine it will be configured to look differently.
Canadian banks buy government bonds.
Nick Rowe: About my crazy voucher idea: I guess if banks cannot lend out the vouchers for interest it would be equivalent to a tax. But what if we are in special circumstances such that interests rates are zero, banks are hoarding cash because they want to build up reserves, and the public also want to reduce debt and save? Then this voucher plans just lets everyone do what they want to do anyway.
What the US is doing now is giving money to banks who hoard it, while homeowners default on their mortgages, so the people go broke while the banks end up with the real assets. If the money went to the people instead of the banks, they would have a better chance of holding on to their real assets.
Inflation is the answer! The debt is nominal, the repayments are real. To reduce the real burden, devalue the currency.
(This works even for expected inflation by the way – because although interest rates rise, so does income. So that part of the repayments that repay the principle shrink in real terms.)
” For the individual, a high debt/equity ratio in the portfolio reduces risk. In aggregate, a high debt/equity ratio increases risk (of financial crises). ”
Conservation of risk – you can’t get rid of it, you can just move it around. High debt/equity ration reduces risk for the individual (huh – do you mean he can only lose his equity because of limited liability – so there is the crux of your paradox – the solvency asymmetry.)
Is the government printing money to redeem its debt (as I think they SHOULD do in a depression, so long as their debts are denominated in their own fiat currency) converting debt to equity or default?
reason:
Inflation: If expected inflation adjusts instantly to actual inflation, and if nominal interest rates on all loans are variable and adjust instantly to expected inflation, then inflation will not affect real values of debts. But as soon as we move to fixed rate debts, inflation will work. The amount it works depends on the average maturity of fixed rate debt.
Conservation of risk: I’m still trying to get my head around this. People want to hold firms’ debt, rather than equity, for safety. So people want a high debt/equity ratio on the ASSET side of their balance sheets to promote safety. But this forces firms to have high debt/equity ratios on the LIABILITY sides of their balance sheets, which is risky. Violates MM of course. Maybe I’m implicitly assuming people are irrational.
printing money: Dunno. If it causes inflation, it’s partially default, and partly an increase in the tax rate on money.
One small point — corporate debt is not remotely as toxic as consumer debt. When individuals assume debts they have no capacity to repay their obligations (plus interest) without borrowing from their future income streams, lowering future demand and savings. On the other hand, a company that borrows money does so (assuming competent management) to create a business that generates revenue from which the original debt plus interest can be repaid.
The business doesn’t need to generate its money from an increase in consumer spending, they can do many things to improve the productivity of manufacturers and earn their money that way. We don’t need new consumer demand to profit, we can make money by doing what we already do, better. That ultimately brings down costs, prices and allows the consumer to both increase savings and demand.
pointbite: i disagree. Consider a home mortgage for example. By borrowing to buy a home you generate a stream of income: the rent you no longer have to pay. And that savings on rent should pay the mortgage. Exactly the same as if a business borrows to build an apartment and rents it out.
Pure consumer loans (not for consumer durables like homes) might be different. But they aren’t necessarily more prone to default, if the future income is there to repay them.
I’m not talking about “default” as much as “productive” use of capital. The fact people are willing to rent their homes from the bank for 30 years before they own it doesn’t mean it’s a good way to deploy capital. However, as you pointed out consumer debt is more than just mortgages, I concede a mortgage is better than borrowing to buy a television, but it’s not equal to business borrowing because the house itself is not generating any revenue. You said, “if future income is there” — that’s my point. In one case we create income from nothing, in the other we take from income that already exists. The income would exist regardless of where you choose to live. And no the rental payment explanation is not good enough, when people buy they tend to move up. The mortgage + appliances + more furniture + upkeep + taxes + utilities + probably more transportation costs + etc are almost always more than their previous rent.
Diverting capital that could have gone to R&D for a house that absorbs much of your disposable income ultimately hurts savings and demand (for anything other than housing and related products/services). Sure you may get a higher standard of living, but at what cost? Is maximizing home ownership the apex of human achievement? That doesn’t mean we should all rent small apartments so capital is cheap enough to fund an army of scientists, but I believe we are out of balance today.
Re printing money – of course it is RELATIVELY inflationary. If deflation is a threat printing money seems to be right solution BECAUSE it is inflationary.
Nick I don’t understand your answer about the effect of inflation on the principal. Am I wrong in thinking that nominal interest rates should adjust to expected inflation to keep real interest rates constant?
Think about this – a bank can make a profit EVEN if real interest rates are negative. They live off the spread between lending and borrowing rates, they don’t worry about the real value of repayments. The negative effect of inflation for banks is on their rent and labour costs.
OK I see what you mean by safety. Equity is last to get paid in the event of insolvency, and dividends are not assured. But this is a common paradox (the paradox of thrift or the tragedy of the commons). What works for an individual doesn’t work for the market as a whole. But this is what the price mechanism is meant to balance. Clearly as you say investors are either myopic or irrational. Is being myopic the same as irrationality?
Nick – re repayment of principal and nominal interest rate adjustment. You are correct it is sufficient as long as there is inflation and not deflation and a liquidity trap. It is DEBT + DEFLATION that is the deadly combination. It is always the asymmetries that kill you.
I should modify – you are right so long as repayments are not made in lump sums. As somebody who has made repayments in lump sums, I’m well aware of the advantage that inflation affords in this case.
I’m now reading Hermann Daly and he has a chapter on Soddy. Maybe Soddy is a least partly right. We need to think harder about the nature of money and the dualism of the financial system and the real economy.
reason:
“Nick I don’t understand your answer about the effect of inflation on the principal. Am I wrong in thinking that nominal interest rates should adjust to expected inflation to keep real interest rates constant?”
That’s what I’m assuming. Real rates stay constant. So inflation has no effect on redistributing wealth between creditors and debtors. Of course, if the nominal interest rate was fixed before expected inflation changes, that can’t happen. It also can’t happen if we get (big) deflation, because nominal interest rates can only adjust so far, down to zero.
Nick,
thanks for replying, but do consider the possible of lump sum repayments. To take an absurd example, imagine if inflation rose to 100% per annum. Then repaying the principle costs more or less the same as paying the interest. Guess what happens.
Good (and VERY LONG) new post from Steve Keen explaining his ideas very clearly:
http://www.debtdeflation.com/blogs/2009/01/31/therovingcavaliersofcredit/