You've probably seen examples. Some poor non-economist says something like: "Governments, like households, must live within their means". And economists all point their fingers and laugh and say that's a fallacy.
So, what are the differences between a government's budget and a household's budget? And do those differences matter? This is just a simple "teaching" post to give you my answers to those questions.
It's not obvious to me that there really are differences that are both qualitatively and quantitatively important.
1 Governments can use force (or the threat of force) to get extra (tax) revenue; households (usually) can't. That difference matters for political questions, and for microeconomic questions about the effect of taxes on the efficient allocation of resources. But it is not obvious why that difference should matter for the question at hand -- about budgets. Doesn't a criminal family that gets its income from mugging also have a budget and have to live within its means? I think the difference is that (most) governments (usually) collect a lot less in taxes than the maximum amount they could collect. (And that is especially true in the short run, because governments can confiscate assets, including refusing to pay their own debts (which amounts to a confiscation), with no outside force to stop them.) So (most) governments (usually) have a choice between cutting their spending to match their revenues and increasing their revenues to match their spending. Now some households will also have that same choice. But perhaps it is rarer for households to be able to increase their incomes to match their spending by the same amount (percentage-wise). But this is a difference of degree only.
2 Governments (usually, or are supposed to) care about the national interest when deciding how much to tax or spend; households (usually, mostly) care about their own self-interest. OK, so governments are like an altruistic criminal family -- like Robin Hood. OK, but doesn't Robin Hood have a budget, and have to live within his means, just like any other household? Yes, but if Robin Hood is altruistic in his collection of revenue (and not just in his giving to the poor) we can understand why a very powerful Robin Hood might usually choose to collect much less revenue than the maximum amount he is able to collect. Which helps us understand my point 1 above.
3 Governments are (usually) much larger than households. Sure, but why does that matter for the question of budgets? It's because of feedback effects. If you are a large fish in a small pond, you will think about how a change in your budget will affect the pond's budget, which in turn will affect your own budget. If you are a small fish in a large pond, that feedback effect will be very small. The Old Keynesian Multiplier is an example of this: if a large government doubles its expenditure, national income will rise, and the government's tax revenue will rise by enough to cover some appreciable portion of the increased spending. If a small household doubles its expenditure, its income will increase by a miniscule percentage, because nearly all the extra national spending will benefit other households, not the original household. But is the Old Keynesian Multiplier analysis right? Well, that depends. It won't work at "full employment", where national income is constrained by the supply-side, and not by the demand-side. And it won't work if the central bank controls the demand-side (as it should if it's an inflation-targeting central bank) and so adjusts monetary policy to fully offset any changes to demand coming from other sources, whether from households or from governments. The whole point of an independent central bank targeting 2% (or whatever) inflation is to do whatever it takes to make the Keynesian Multiplier precisely zero. So there won't be any feedback effects.
4 Governments can print money; households (usually) can't. OK, now we are getting to the qualitative differences; or are we? And does that qualitative difference matter quantitatively?
First off, many governments (like Canadian provincial governments, and Eurozone national governments) can't print money. And some households (those which own banks, or bank shares) can print money (except it's "printed" electronically, which doesn't matter a jot).
But there is nevertheless a difference between money printed by the central bank and money printed by commercial banks. Actually, there are two differences:
The first difference is that the central bank (usually) has a (de jure or de facto) monopoly in printing currency; commercial banks (usually) face more competition from other commercial banks in "printing" demand deposits. And monopolies can (usually) earn monopoly profits (which get given to the government if the government owns the central bank). Competitive firms, disciplined by entry of new firms, (usually) can't.
The second difference is that commercial banks (usually) peg the exchange rate of their money to the money issued by the central bank (they promise to redeem at par); the central bank does not peg the exchange rate of its money to the moneys issued by the commercial banks. The central bank is the alpha leader, that can choose to make its money more valuable or less valuable; the commercial banks are the beta followers, that just follow the central bank in making their moneys more valuable or less valuable. So the government can always order the central bank to inflate away the value of its money, and so inflate away the value of government debts which promise to pay fixed amounts of that money; and the commercial banks just follow along. Commercial banks (and the households that own them) can't do this.
Sure, that's a qualitative difference between a government's budget and a household's budget. But does it matter in practice? And does it matter quantitatively in practice? No, not if the central bank sticks to its 2% inflation target, because that means it will not exercise its power to inflate away (a de facto confiscation) of the government's debts. That 2% inflation target has presumably already been factored into expected inflation and the nominal interest rate paid on government (and household) debt. And the government's revenue from the central bank's monopoly profits from issuing currency aren't that big either. For a back-of-the-envelope calculation: assume a 5% currency/GDP ratio, and a 5% growth rate in Nominal GDP (3% real plus 2% inflation), so 5% x 5% = 0.25% which means the government revenue from printing money is 0.25% of GDP. A nice little Magic Money Tree of monopoly profits, but no biggie. Other sources of government revenue are much bigger.
But in an extreme emergency, when the 2% inflation target can be cast aside? Maybe. But the government could also just confiscate stuff.
5 Governments have (potentially) infinite lives; households have finite lives. That's what economists usually say, but is it true? And why should it matter?
Governments sometimes fall to revolutions which restart everything from scratch, and nations can disappear. And people have kids, and grandkids, or can "adopt" them if they want to continue their business after they die. Or set up trusts and corporations. What's the difference?
Let me cut to the chase: ignore 5, and let's go straight to 6:
6 Governments can (usually) bequeath net liabilities; households (usually) can't. I can bequeath my assets to my kids, and I can also bequeath some of my liabilities to my kids, like a house and the mortgage that goes with it. But my kids can always refuse to accept my bequest, and they presumably will refuse to accept my bequest if they figure the assets are worth less than the liabilities. I can't make my kids accept a negative bequest. (Though commenters on an old post once told me it used to be the law in some countries, and maybe still is in others, that kids are responsible for their parents' debts.) But governments can force my kids (as future taxpayers) to accept a negative bequest of the national debt. (Though even then, my kids might emigrate if they figure it's too negative.)
Economists who say "the national debt is not a burden on future generations of taxpayers, because they inherit both the bonds and the tax liabilities inherent in those bonds, and so owe it to themselves" don't get this. They are playing fast and loose with the word "inherit". Now it's true that if the kids literally do "inherit" the bonds from their parents, as a freebie, then there is no net negative bequest (distortions from taxation to finance the bonds aside). Because the positive bequest from their parents' bonds offsets the negative bequest from the government debt. But if instead we sell the kids the bonds, they are paying us for an asset which is their own liability. It's like buying an IOU, and then seeing your own signature on that IOU. They have inherited a liability, which makes them worse off, unless they can palm it off to the grandkids in turn.
Which is where things get weird. Because if the interest rate on government bonds is always less than the growth rate in the economy, it is possible for them to palm it off to the grandkids, and the great-grandkids, forever. It's a Ponzi scheme, where you borrow more to pay the interest on the debt, and so the debt grows at the rate of interest. But unlike the real Mr Ponzi's scheme, there is no reason it should ever end, provided the economy always grows faster than the debt. You can actually borrow more than is needed to pay the interest on the debt, to keep the debt/GDP ratio from falling over time. And some government bonds do in fact (usually) have interest rates lower than their growth rates.
But why can't a dynastic household do the same thing? "OK kids, you get to inherit my debts. But at the same time you inherit my name and reputation that keeps the interest rate on my debts below the growth rate of the economy, so you can make a nice living from issuing more debt to cover your reasonable living expenses." Dunno. Maybe they can, if the stars line up right, and they never break the chain.
I think I will stop there.
interesting
might be augmented by comparing government, corporate, and (unincorporated) household budgets
Posted by: JKH | October 28, 2017 at 12:58 AM
interesting
might be augmented by comparing government, corporate, and (unincorporated) household budgets
Posted by: JKH | October 28, 2017 at 12:58 AM
Government debt works on longer timescales than household debt. Although this is not an "in principle" difference, it has a large effect in practice. If a household runs up their credit card debt, calls in favours from family and friends, pawns off their valuables... they might last a year or two before it all catches up with them.
On the other hand if you look at the Greek government entering the EU, they managed to keep rolling over increasing debt for something like 20 years before they went broke, and even then you still have people arguing that Greece is not broke.
So in terms of perception, and fooling some of the people some of the time, that long debt buildup can convince people that everything it going to be OK. It feels different to a household, even when you might argue it is not entirely different.
Posted by: Tel | October 28, 2017 at 01:04 AM
JKH: thanks. Is the main difference bankruptcy law? Corporations automatically default if their net assets are negative, so their shareholders aren't liable?
Posted by: Nick Rowe | October 28, 2017 at 07:15 AM
Tel: you may have a point there. Mortgages may be an exception, because they can be very long term. But maybe mortgages are different, because they are secured by an even longer term asset?
Posted by: Nick Rowe | October 28, 2017 at 07:18 AM
Nick,
I was thinking also of the nominal actual/potential lifetime of a long-lived corporation compared to that of a household. Maybe that's a superficial distinction, along the lines of your # 5. Although it seems to me that a corporation lies between government and households in some sense of the characteristics under discussion.
Posted by: JKH | October 28, 2017 at 08:01 AM
As part of his overall argument that differences between the public and private sector don’t matter, Nick says in reference to central banks and commercial banks respectively, “And monopolies can (usually) earn monopoly profits (which get given to the government if the government owns the central bank). Competitive firms, disciplined by entry of new firms, (usually) can't.”
The suggestion there is presumably that money creation by private banks doesn’t matter. I suggest that actually it does matter, or “do harm”, and for the following reason.
If an economy uses just (or mainly) base money, and private banks create and lend out more money, that will be inflationary, assuming the economy is already at capacity, which means the state will have to impose some sort of deflationary measure, like confiscating base money from the private sector via tax. So essentially commercial banks steal from the rest of the private sector in just the same way as backstreet counterfeiters do. In fact the French Nobel laureate economist, Maurice Allais regarded private money creation by private banks as being essentially counterfeiting.
Put another way, money creation by commercial banks enables that seigniorage to subsidise bank loans, as Joseph Huber explained on p.31 of his work “Creating New Money” – see para starting “Allowing banks..”. See:
http://www.jamesrobertson.com/book/creatingnewmoney.pdf
The net result is artificially low interest rates and an artificially large amount of lending, i.e. debt.
I expanded on the counterfeiting point recently here:
https://medium.com/@ralph_47183/most-money-is-counterfeit-725c1f7f98c6
Posted by: Ralph Musgrave | October 28, 2017 at 12:10 PM
Hi Nick,
Thanks for this--very insightful. As I reflect, some of the stories about why government is different are about why government can borrow more cheaply. Since govt can raise resources (through inflation 'tax' or explicit tax) to pay off debt in future, govt is a better risk than Jane Household who does not have those tools. This just means govt gets a better deal (lower i; better terms) for its debt than households.
Is 'better price' an *essential* difference? Perhaps; perhaps not. But I agree with you that a *bigger* essential difference is the time horizon, and specifically your insight that debts are bequeathed.
This is highlighted by the fact that the specific example that seems to sort laypeople and economists is the idea that we never have to pay back debt, which is what underlies the focus on debt/GDP. I have found that this notion stokes the fire of some people because they see it as inherently ridiculous. "Of course it is a burden on our children!" they say. In contrast, the economist sees it is eminently reasonable since we know governments roll over debt all the time.
Specifically, the response on twitter to my claim that a small deficit can be sustained to infinity without debt explosion (which is true!) was met with incredulity. This example highlights Nick's point well--the key thing that makes government different is the long-term/bequeathable nature of debt.
So, I think Nick is right about this; but I think the idea can be pushed even further to think about *why* debt is bequeathable.
On twitter, I pushed the example of Russia which accepted some responsibility for *Tstarist* debt http://www.independent.co.uk/money/eighty-years-later-the-tsarist-bond-pays-off-1283278.html
I also mentioned the example of 16th century Spain, which kept being able to borrow in spite of serial defaults https://press.princeton.edu/titles/10084.html.
Also, modern Argentina, which I'm told by colleague Mauricio Drelichman keeps being able to sell bonds at terms that exceed the maximum span of any constitution in its history.
These examples make clear Nick's point: lenders are willing to lend to government because debts can be bequeathed--even across political/constitutional regimes.
Let me push Nick's conjecture even further. The reason for this debt bequeathing is place. No matter what happens in the world, we pretty much know where to find Russia on the map. Yes, the periphery changes across centuries, but I'm pretty confident I can find "Russians" in the 22nd century if I have a bond to cash. I don't care what government of what type is running Russia at that time. I just know that if "Russians" want to borrow internationally again, they have to cash my bond.
(Although we could push this further--what if we destroy earth and are on an Ark headed to a colony on Ceti Alpha 4? Would we try to find "Russia" on the ark in order to get paid? Maybe it is the persistence of perceptions of ethnicity more than 'place' which leads to bequestable debt....)
So, I think Nick has nailed the essential difference--bequestable debt. I have a conjecture about how to dig deeper--it is place/ethnicity that drives *why* government debt is seen by markets as bequeathable.
Posted by: Kevin Milligan | October 28, 2017 at 01:00 PM
Ideally a (national) government shouldn't be an entity separate from the national economy. The austerity argument is usually made with the conceptual background that the government extract money from the economy, and this money then disappears. But what really makes money disappear from the "real" (goods and services) economy is economic rents siphoned off and cycled into a financial machinery where it mostly rotates between investment papers. The government doesn't hold on to most of its revenue but pays it back to various economic activities.
Governments also have other means (legislation) to control money flows in the economy - the money doesn't have to go through the government by taxation.
Seeing how the argument is usually made, it means the government should not fund public services but direct the money to connected businesses while taxing everybody else.
That's not to say there is no waste in the public sector - but I would like to see a few larger private sector businesses where there is no waste.
Posted by: cm | October 28, 2017 at 03:13 PM
AFAICT there is no growth limit on bond issuance. You can have an economy that has a resolutely zero GDP and a positive interest rate andthe bond system appears to work just fine
Always remember currency is just a zero interest bearer bond from the perspective of anybody outside that currency zone and bonds are always denominated in the currency of a currency zone.
Posted by: Neil Wilson | October 28, 2017 at 04:00 PM
JKH: IIRC, universities and churches, which are corporations (though not the usual sort), sometimes tend to be very long-lived. If my monetary history were better, I would also talk about medieval banking families, but my memory is too hazy.
Ralph: maybe think of it this way: commercial banks offer a product (chequeing accounts) that is a substitute (though not a perfect substitute) for government money (currency). Whenever a substitute exists, it reduces the demand for the monopolist's product, and reduces the profits the monopolist can make. So yes, if commercial banks were banned (or made to hold 100% currency reserves), the demand for currency would increase, so the government-owned central bank would earn more revenue from printing money. If, for example, it quadrupled the demand for currency, my back-of-the-envelope calculation says that the government now earns 1% of GDP from printing money compared to 0.25% (assuming we hold the inflation target constant).
Kevin: thanks! I think it's really important to distinguish between r > g and r < g. With r > g, the Long Run Government Budget Constraint is well-defined as T goes to infinity. The PV of primary surpluses must equal the current debt. In a steady state, with debt/GDP ratio constant, there's a primary surplus, but a deficit as normally defined. It's the Samuelsonian r < g case that is genuinely weird, because the Long Run Government Budget constraint is not well-defined. It does not converge. So I have a lot of sympathy with non-economists not getting this. But even in this case, I think we need to remember that demand curves for government bonds (and money) slope down, so that r will be an increasing function of the debt/GDP ratio. So there is a limit to how much of a primary deficit the government can run, even in the r < g case. (Just like there's a revenue-maximising inflation rate in the case of printing currency, which pays 0% nominal interest and negative real interest.)
Posted by: Nick Rowe | October 28, 2017 at 05:42 PM
Neil: in an economy with infinitely-lived households (dynasties where the kids inherit the bonds) and lump-sum taxes there is indeed no limit to bond issuance. The stock of bonds is irrelevant. This is Barro's famous Ricardian Equivalence result "Are bonds Net Wealth?". We owe it to ourselves. (But then fiscal policy in the form of bond-financed tax cuts has zero effect also.) But with finite-lived agents, who *sell* their bonds to their kids, the stock of bonds does matter, and there is a limit. Very simply, if the debt/GDP ratio is too high, the young generation will not have enough income to be able to buy the bonds from the old.
Posted by: Nick Rowe | October 28, 2017 at 05:57 PM
"The whole point of an independent central bank targeting 2% (or whatever) inflation is to do whatever it takes to make the Keynesian Multiplier precisely zero. So there won't be any feedback effects."
So does a government have a role in setting macroeconomic targets and performance regarding output and employment when the economy is below full employment? Is it all to be left to the CB?
Posted by: Henry | October 28, 2017 at 06:47 PM
Nick: "Mortgages may be an exception, because they can be very long term."
I don't think it's so very different. I'm looking at the key phrase "must live within their means", thus the critical metric is the time delay between making a spending decision that blows your budget and experience of the consequence of that decision.
Although a mortgage is long term debt, if you miss payments on that mortgage the bank will certainly not sit back and wait a long time before they start to push you. Thus, the link between cause and effect is not long term. I think the bank will be chasing you within perhaps 3 months to 6 months, and they will have the lawyers involved within 12 months if you keep missing payments.
So governments are capable of significantly longer delay of the consequences of their spending decisions than households. That's what I'm getting at. I could go a step further and point out that in a household the decision makers themselves will personally see the consequences within their own lifetimes... while in government the individual decision makers will be retired before any consequences arrive. This does not mean there are no consequences, or somehow those don't matter, not at all. But it does allow plenty of obfuscation in the middle.
Posted by: Tel | October 28, 2017 at 07:56 PM
To be fair to Paul Krugman, he actually has a different argument, not represented on the list above. That's not to say I agree with him, and I don't believe he would be fair to me if the shoe was on the other foot, but here goes:
https://krugman.blogs.nytimes.com/2013/03/14/running-government-like-a-business-or-family/
For what it's worth, I sincerely hope that none of my spending ever becomes Paul Krugman's income.
Posted by: Tel | October 28, 2017 at 08:12 PM
Henry: that's the idea behind central banks being responsible for targeting inflation, yes.
Tel: in the olden days, rich altruistic households might sometimes increase their spending to offset recessions in their local communities. (The Hellfire Caves are supposed to be an example.) I think Paul Krugman's point there is covered in my 2 and 3.
Posted by: Nick Rowe | October 28, 2017 at 08:32 PM
Another case of making what is false in theory, true in fact, as this is entirely dependent on government doing the right thing when what is being suggested is doing the wrong thing, that is, there are times when budget constraints matter, and times when they do not, or more accurately, matter in precisely the opposite direction.
Posted by: Lord | October 28, 2017 at 08:53 PM
Nick: in point 3, if I understand correctly, you dismiss the case for active fiscal policy: "And it won't work if the central bank controls the demand-side (as it should if it's an inflation-targeting central bank) and so adjusts monetary policy to fully offset any changes to demand coming from other sources, whether from households or from governments. The whole point of an independent central bank targeting 2% (or whatever) inflation is to do whatever it takes to make the Keynesian Multiplier precisely zero. So there won't be any feedback effects."
Doesn't the experience of Japan since the early 1990s, and then the Great Recession, tell us that monetary policy is indeed ineffective at the zero bound (we can't lower interest rates past zero), and that there's a case for fiscal policy as a second-best solution? Maybe in theory there's a way to use monetary policy even at the zero bound (e.g. NGDP targeting), but we don't have any real-world experience here.
Common sense and intuition tells us that when times are tough, everyone needs to cut back spending - households, firms, governments. But the point of Krugman's argument is that cutting back public spending to balance the budget during a recession is self-defeating. And it's natural for deficits to rise during a recession, because of automatic stabilizers like unemployment insurance and falling tax revenue on profits.
To take the other side of the argument, another point against feedback is that in a small open economy like Canada's, the impact of looser fiscal policy will be spread out over one's trading partners. So without coordinated action by multiple governments, looser fiscal policy won't be that helpful. Kevin Milligan.
Posted by: Russil Wvong | October 29, 2017 at 02:50 PM
Russil: there's a debate about whether monetary policy can work (or whether help from fiscal policy is needed) when nominal interest rates are very low. I think that monetary policy still works, but I want to stay out of that debate for this post. But yes, *if* you think that monetary policy is unable to keep spending on track, so the central bank will not in fact offset fiscal policy, then my points 2 (altruism) and 3 (large player) can be used to make the standard case for fiscal policy. (And you are also right about the problem of fiscal policy in a small open economy).
Posted by: Nick Rowe | October 29, 2017 at 03:01 PM
>> But is the Old Keynesian Multiplier analysis right? Well, that depends. It won't work at "full employment", where national income is constrained by the supply-side, and not by the demand-side.
There's also supply-side feedback that operates on longer timescales. Canada's debt-to-GDP ratio would probably get worse and not better if it reduced the current deficit by cancelling funding for police and court services.
At the household level, this is equivalent to the less-controversial idea of going into debt to acquire human capital or real capital, however in some Internet commentary I see occasional invocations of the idea that the government itself should refuse to go into gross debt to buy roads or hospitals.
Posted by: Majromax | October 29, 2017 at 03:09 PM
Nick:"Mortgages may be an exception, because they can be very long term. But maybe mortgages are different, because they are secured by an even longer term asset?"
Fat lot of good it did to the Spanish economy. No runaway gunmint profligacy, all seriously backed private investments. Trichet and Schaüble still managed to wreck the country. And us usual, it may lead to the normal conclusion of economic mismanagemnt, civil-ethnic war.
Posted by: Jacques René Giguère | October 29, 2017 at 03:19 PM
Majro: I agree. Neither a government nor a household that borrows to finance a profitable investment is "living beyond its means". They are equivalent in that respect.
Jacques Rene: I see the Spanish economy as like a Canadian province, only with a worse monetary policy from the ECB than the Bank of Canada.
Posted by: Nick Rowe | October 29, 2017 at 04:22 PM
Nick,
7. The legal protections given to holders of sovereign government liabilities are more stringent than those given to holders of household liabilities.
Households can declare bankruptcy in a court of law and receive relief from creditors. Sovereign governments cannot.
This amounts to the flip side or your #1 - because households cannot forcibly increase their tax revenue within a legal framework while the federal government can forcibly increase it's income stream (tax revenue) again within a legal framework, households can be relieved of their debts in a bankruptcy proceeding.
Posted by: Frank Restly | October 29, 2017 at 05:02 PM
Sorry, number 4 nails it. Most households don't inflate to pay for stuff, but governments sometimes do. and I noticed, sadly, a while back that money creation wasn't part of my budget constraint.
Durn.
Posted by: Pete Bias | October 29, 2017 at 07:42 PM
"The second difference is that commercial banks (usually) peg the exchange rate of their money to the money issued by the central bank (they promise to redeem at par); the central bank does not peg the exchange rate of its money to the moneys issued by the commercial banks. The central bank is the alpha leader, that can choose to make its money more valuable or less valuable; the commercial banks are the beta followers, that just follow the central bank in making their moneys more valuable or less valuable."
I do not agree with this at all. As long as the commercial bank(s) is/are solvent, I would say the central bank does peg/fix the exchange rate of its central bank money to the moneys issued by the commercial banks. The commercial banks are also alpha's.
The U.S. is not on a "CPI standard". The U.S. is on a "demand deposit of the solvent commercial banks standard" with an indirect price inflation target.
Posted by: Too Much Fed | October 29, 2017 at 09:34 PM
One additional thing to consider is if the CB is constrained in its use of particular monetary instruments (purchase of ST government debt) then it's policy may not be independent of the deficit. [This is my interpretation of Bernanke's plea for "help" from fiscal policy in 2009.] Additionally, governments may (because of past failures to invest as much as would be indicated bu a NPV rule (="austerity") have investment opportunities that yield positive returns when discounted at the CB lending rate. This may be especially the case if the opportunity costs of inputs into these investment opportunities during a recession are less than the market prices, that is if there are unemployed resources. Both these may make a government less credit constrained than a household.
Posted by: Thaomas | October 29, 2017 at 10:16 PM
" r will be an increasing function of the debt/GDP ratio"
The g and the r are the long term rates, and r is real, right? So isn't r just the subjective discount rate + (risk aversion * g)? I.e. constant?
Posted by: notsneaky | October 30, 2017 at 01:58 AM
notsneaky: we can either interpret r and g as both real, or both nominal, it should work out the same either way.
"So isn't r just the subjective discount rate + (risk aversion * g)? I.e. constant?"
That will only be true for the Euler equation in an infinitely-lived representative agent model, where Ricardian equivalence holds, so debt is irrelevant. It won't work in an overlapping generations model, where real r will be an increasing function of the debt/GDP ratio.
Posted by: Nick Rowe | October 30, 2017 at 06:04 AM
Related to 4: Governments can establish currency. If it decided to only except tax payments in ripe peaches, well, that would be the new currency. Households must take what exists.
Related to 6: it always bequeaths the largest national asset of all, the PV of future tax receipts plus tax optionality.
Posted by: john | October 30, 2017 at 09:10 AM
john: " If it decided to only except tax payments in ripe peaches, well, that would be the new currency."
No it wouldn't. Peaches would make a rotten currency. People would use something else as currency, then buy peaches to give to the government at tax time.
Take a real-world example: governments often collect taxes in the form of labour: a military draft is an example. But that doesn't mean people suddenly switch to using labour as a medium of exchange or unit of account.
Posted by: Nick Rowe | October 30, 2017 at 10:18 AM
This post makes the Venezuelan people feel much better about their situation.
Posted by: Jay | October 30, 2017 at 09:52 PM
Jay: communism doesn't work. Which is why Venezuela is a mess. But this post is not about communism.
Issuing too much debt or money is also a bad thing. Which is why it is important to understand how much is too much.
Posted by: Nick Rowe | October 31, 2017 at 06:16 AM
"It won't work in an overlapping generations model, where real r will be an increasing function of the debt/GDP ratio" - even in steady state? What pins down the capital stock?
Posted by: notsneaky | October 31, 2017 at 09:57 AM
"Take a real-world example: governments often collect taxes in the form of labour: a military draft is an example. But that doesn't mean people suddenly switch to using labour as a medium of exchange or unit of account."
Maybe you can't do so now, but elsewhere elsewhen you could hire a guy to take your spot in an army. The draft example is not good because you still have another official currency. Peaches is meant to be ridiculous, but the only currency. Use marbles instead if you like.
Now, you may remember Paul Krugman's discussion of the liquidity crisis among the members in a day care coop. They accumulated chits for units of labor. The solution to the crisis was to print chits. Labor chits was the only currency.
"People would use something else as currency..." suggests finding an example.
Posted by: john | October 31, 2017 at 10:56 AM
notsneaky: Even in steady state.
The simplest model is a 2-cohort overlapping model with no capital. With log preferences, 1+r=C(old)/C(young) (ignoring pure subjective time-preference-impatience). Every period, the young buy debt from the old. The bigger the debt, the smaller is C(young) and the bigger is C(old).
We can add kapital to the model, but if there are diminishing returns, the K/L ratio is determinate.
Posted by: Nick Rowe | October 31, 2017 at 01:53 PM
What is K/L ratio determined by? In the Ramsey model you have r=MPK=d+theta*g and you invert that to get steady state k. Why would it be different in OLG? (And if I'm not mistaken in steady state (1+r)=1/discount factor, so you can't ignore pure subjective time preference)
Posted by: notsneaky | November 01, 2017 at 09:11 AM
I mean, I think in any dynamic model with capital which has a steady state you gonna have r=impatience etc. in steady state, no? Otherwise it wouldn't have a steady state. So impatience determines r, which determines MPK, which determines k. And k would also determine debt/gdp ratio, I think, and r is exogenous.
Posted by: notsneaky | November 01, 2017 at 09:15 AM
notsneaky: It's more complicated in an OLG model than in a Ramsey model. Because it's simultaneous.
Remember that even if aggregate per capita C is constant over time, this does not mean that an individual agent's C is constant over his lifetime. Because he saves when young, to buy K+B from the old, and dissaves when old, when he sells K+B to the next generation of young. So r must equal impatience + (some elasticity times)the growth rate of an individual's consumption over his lifetime. The simplest model is where they produce when young, so consume Y-B-K, and retire when old, so consume (K+B)(1+r). )
Posted by: Nick Rowe | November 01, 2017 at 11:48 AM
"Economists who say "the national debt is not a burden on future generations of taxpayers, because they inherit both the bonds and the tax liabilities inherent in those bonds, and so owe it to themselves" don't get this. They are playing fast and loose with the word "inherit". Now it's true that if the kids literally do "inherit" the bonds from their parents, as a freebie, then there is no net negative bequest (distortions from taxation to finance the bonds aside). Because the positive bequest from their parents' bonds offsets the negative bequest from the government debt. But if instead we sell the kids the bonds, they are paying us for an asset which is their own liability. It's like buying an IOU, and then seeing your own signature on that IOU. They have inherited a liability, which makes them worse off, unless they can palm it off to the grandkids in turn."
Come on, if that was true the bonds would not have any value. Who's buying an IOU with her or his own signature on it? Are you saying that any corporate debt is similarly a burden on future generations? Why not? Or stocks? Any financial asset would do! Or what *are* the differences between government debt and corporate debt?
Posted by: Jussi | November 02, 2017 at 08:20 AM
Jussi: If the government can legally "forge" my signature on an IOU, and make it enforceable, I will have to pay that IOU whether I buy it or whether someone else buys it. Sure, the kids could all get together and collude as monopsonists to only pay their parents one penny on the dollar for the bonds, but that doesn't happen.
Yes, corporate bonds are a liability of future generations of shareholders. But they can and will refuse to accept those liabilities if the assets of the corporation are worth less than the liabilities.
Posted by: Nick Rowe | November 02, 2017 at 11:45 AM
I'm not sure what you mean by "forge". The government will sell bonds just like a corporate. The government debt is then a liability for the tax payers just like the corporate debt is a liability for the corporate owners. And all financial agreements are enforceable. Would you also say that CFO/CEO can legally "forge" shareholders' signature on the corporate IOU?
Do you mean "not buy" by "refusing to accept"? Then this applies to the government bonds too without any collusion. (They will not buy those liabilities if the assets of the sovereign are worth less than the liabilities). If you mean something else, please elaborate.
Posted by: Jussi | November 02, 2017 at 03:11 PM
Posted by: Tel | November 03, 2017 at 06:18 AM
Nick,
Nice piece. I suspect the thinking dates back to an era - not so very long ago - when consumer credit, outside of mortgages or car loans, was unusual. In an era before credit cards for all unsecured lines of credit, and pay day loans, it was probably true that households had to live within their means and governments didn't. With the rapid expansion of consumer credit over the past 3 decades, that's no longer clear.
Posted by: Bob Smith | November 03, 2017 at 08:23 AM
Nick, did you mean the same by "refuse to accept" as Tel is likely implying? Somehow being able to default your debt obligation would make "the kids" better off in the corporate case? I do not get, defaulting is just a way to distribute wealth from creditors to borrowers. It doesn't change the aggregate.
So I ask again: what *are* the differences between government debt and corporate debt in terms of being burden on future generations?
Tel, please elaborate?
Posted by: Jussi | November 03, 2017 at 08:26 AM
Jussi,
"It doesn't change the aggregate."
It can change the aggregate going forward.
How many times would you keep lending to a corporation if they kept defaulting on their loans?
Posted by: Frank Restly | November 03, 2017 at 10:17 AM
If you already understand what I'm implying, then there would not appear to be much left to elaborate on.
The corporate debt is NOT a liability for the corporate owners, because they are shareholders in a Pty Limited corporation.
If no one wants to become a new shareholder, they simply choose not to buy shares, and the price crashes to zero. Nick got that one exactly right.
Posted by: Tel | November 04, 2017 at 12:21 AM
"The corporate debt is NOT a liability for the corporate owners, because they are shareholders in a Pty Limited corporation."
Huh, whose liability is it then?!
"If no one wants to become a new shareholder, they simply choose not to buy shares, and the price crashes to zero. Nick got that one exactly right."
Tel, I already argued that this doesn't change the aggregate. You don't create wealth by defaulting. It just changes the distribution. Isn't that quite obvious? I don't believe that Nick meant that?
Posted by: Jussi | November 04, 2017 at 06:50 AM
"The corporate debt is NOT a liability for the corporate owners, because they are shareholders in a Pty Limited corporation."
Huh, whose liability is it then?!"
Jussi, I believe corporate debt is a liability of the corporation itself.
Posted by: Too Much Fed | November 04, 2017 at 12:40 PM
TMF and Tel(?),
Yes, if you are a lawyer. No, if you are an economist. Owners of a corporation are liable up to the corporate's assets (for a limited corporation anyway) and if assets do not cover the liabilities it's the creditors who bear the losses.
The question here is why "refuse to accept" makes the kids better off. Maybe you are thinking that creditors are not part of the kids cohort? So are you saying that if all corporations would be "unlimited" corporate debts would also be a burden? Why one type of debt is a burden on future generation but another is not?
Or think about it this way:
There cannot be a financial asset without the corresponding liability. Those two always coexist! A financial asset is just an agreement how to deal with current or future real goods or services. If one gets more someone else gets exactly as much less. One man's debt is another man's asset. Net of financial agreements is always zero, all the time. So there cannot be any kind of time travelling (e.g. http://worthwhile.typepad.com/worthwhile_canadian_initi/2015/02/debt-does-have-intergenerational-distributional-implications.html). IMO, this seems rather simply, so I don't understand why some smart economists like Nick and many others are saying that there can a burden (through debt = financial liability) without a corresponding (financial) asset?
Posted by: Jussi | November 04, 2017 at 03:32 PM
Jussi,
"One man's debt is another man's asset. Net of financial agreements is always zero, all the time."
But the only place where pure financial agreements are reached is in the government sector. Holding U. S. government debt does not grant you a claim on the physical property owned by the federal government nor does it grant voting rights in excess of what are available to any other citizen (Holding $10 million of government debt does not give me 10 million votes). Instead government debt is a pure financial transaction, I trade the government some dollars now for dollars later paid from either tax revenue / additional borrowing.
As you mention, In the private sector voting rights / transfer of real goods / services are attached to many contracts (debt / equity).
If I don't make the payments on my mortgage, the bank takes my house.
When I buy shares in company XYZ, I receive voting rights in hiring / firing decisions.
What you should realize is that the supply of those real goods can expand / contract in such a way that makes both parties in a contract winners or both parties in a contract losers.
"If one gets more someone else gets exactly as much less."
In terms of private contracts, one man's loss is not always another man's gain. Both men can lose under certain conditions.
Posted by: Frank Restly | November 04, 2017 at 05:19 PM
Frank,
"In terms of private contracts, one man's loss is not always another man's gain. Both men can lose under certain conditions."
This is obviously true, but that wasn't the original argument, was it? Nick wasn't saying that government debt is bad because of crowding out, inflation, suppressed exports, etc. That is a totally different kettle of fish. So that's a discussion for another day.
But it's good that you brought up mortgages! Those liabilities are not limited to a certain set of assets - they might eat them all! So (practically) the household debt is a burden along with the government debt? All of those evil credit cards? And so are by the way our beloved notes and coins?
Come on.
Posted by: Jussi | November 04, 2017 at 07:27 PM
Jussi,
I am trying to answer your original question (November 2, 2017) - "Or what *are* the differences between government debt and corporate debt?".
With corporate debt (or mortgages), the transfer of real resources can be a condition of the debt contract.
With government debt (these days anyway) that is rarely if ever the case.
And again (November 3, 2017) - "So I ask again: what *are* the differences between government debt and corporate debt in terms of being burden on future generations?"
As far as a burden on future generations, I as a citizen bear the burden of taxation and thus am liable for the interest payments on the sovereign debt (including the accumulated debt). I bear no such burden to buy a Ford truck or a Sony television in helping these corporations pay off their debt.
And finally again (November 4, 2017) - "IMO, this seems rather simply, so I don't understand why some smart economists like Nick and many others are saying that there can a burden (through debt = financial liability) without a corresponding (financial) asset?"
One answer is that the central bank can make loans and retain those loans. This creates a liability in the private sector with no corresponding asset (bond).
Another (alluding to what I have already said) is that the value of the asset and liability can change independently of each other.
This is especially true when the transfer of real goods is included in the contract.
Posted by: Frank Restly | November 04, 2017 at 08:03 PM
"Yes, if you are a lawyer. No, if you are an economist. Owners of a corporation are liable up to the corporate's assets (for a limited corporation anyway) and if assets do not cover the liabilities it's the creditors who bear the losses."
I believe you are describing a reduction in corporate equity here. I would not call that liable.
"One man's debt is another man's asset. Net of financial agreements is always zero, all the time. So there cannot be any kind of time travelling"
If I am reading that right and a bank is involved, I do not agree with that at all.
Posted by: Too Much Fed | November 04, 2017 at 11:30 PM
Frank,
Nick argument: "It's like buying an IOU, and then seeing your own signature on that IOU. They have inherited a liability, which makes them worse off, unless they can palm it off to the grandkids in turn."
So it is not about real resources directly. That's another discussion.
"As far as a burden on future generations, I as a citizen bear the burden of taxation and thus am liable for the interest payments on the sovereign debt (including the accumulated debt). I bear no such burden to buy a Ford truck or a Sony television in helping these corporations pay off their debt."
Some people ("you") pay taxes to pay off debt and some not. Similarly, some people are shareholders and some ("I") are not. So it is a distributional question. No, it doesn't change the aggregate.
"One answer is that the central bank can make loans and retain those loans."
It doesn't work like that (for an economist). If true the government could create infinite wealth (printing assets without corresponding liabilities).
"Another (alluding to what I have already said) is that the value of the asset and liability can change independently of each other."
No they cannot (yes, if you are an accountant).
Posted by: Jussi | November 05, 2017 at 03:53 AM
TMF,
"I believe you are describing a reduction in corporate equity here. I would not call that liable."
Yes, corporate debt means lower shareholders' equity value. So that's a burden for the shareholders, regardless what you call it.
"If I am reading that right and a bank is involved, I do not agree with that at all."
Give us an example! But keep in mind that it is easy to create financial assets, if their net value is not zero it would be easy to create infinite amount of wealth.
Posted by: Jussi | November 05, 2017 at 03:57 AM
Jussi: "Give us an example! But keep in mind that it is easy to create financial assets, if their net value is not zero it would be easy to create infinite amount of wealth."
TMF is right. See my old post. But it does not follow that you can create an infinite amount of wealth by creating financial assets, just like you cannot create an infinite amount of wealth my creating any individual good. Because demand curves slope down, so if you create too much the price drops below the costs.
And Tel is right that the shareholders of a corporation can and will walk away from the liabilities of that corporation, if the assets are less than the liabilities.
Posted by: Nick Rowe | November 05, 2017 at 07:21 AM
Nick, the old post is about money and liquidity. How is it relevant to this question? Are you saying that the government debt is a burden and other type of debt is not because money might have some value due to the network effect?
"And Tel is right that the shareholders of a corporation can and will walk away from the liabilities of that corporation, if the assets are less than the liabilities."
Nick, no one said they couldn't. But again is this the reason why government debt is a burden and corporate debt is not? Are you saying that "walking away" will magically create value for the kids? Where does that wealth come from? Are you REALLY saying that whether or not a corporate is a limited company will have intergenerational consequences?!
Posted by: Jussi | November 05, 2017 at 08:43 AM
Jussi: "Give us an example! But keep in mind that it is easy to create financial assets, if their net value is not zero it would be easy to create infinite amount of wealth."
I think Nick is agreeing with me for a different reason.
I save $10,000 in demand deposits. I start a bank by selling $10,000 in stock to myself. The reserve requirement is 10%. The capital requirement is 10%. The bank creates $100,000 in demand deposits (creating a reserve requirement of $10,000). Borrowers create $100,000 in new bonds. There is an asset swap of demand deposits for bonds. The bonds purchased by the bank now create a capital requirement of $10,000. The demand deposits are spent in the real economy.
There is a saving flow of $10,000. There is a dissaving flow of $100,000. NGDP (a flow) increases by $90,000 in the present.
My saving decreases NGDP by $10,000 in the present. When I spend it, NGDP increases by $10,000. So, I consider saving spent to be "past" "money" brought to the present to increase demand.
The dissaving increases NGDP by $100,000 in the present. So, I consider dissaving spent to be "future" "money" brought to the present to increase demand.
The "money" is time traveling around to different time periods (most likely) relative to what would happen if there was no saving and dissaving.
Finally, nobody is probably going to agree with me about that.
Posted by: Too Much Fed | November 05, 2017 at 03:22 PM
I can see most of the "monthly budgeting" discussed here is about the currency denominated debt.
I would say the right questions are:
1) Is the debt/bonds bought by a bank or non-bank?
2) What entity is supposed to make the principal payments and interest payments and when?
3) What happens if there is a default?
In general, gov't bonds and household bonds can be bought by a bank or non-bank.
In general, household bonds have dates when principal and interest payments are supposed to be made. Most of the time, the gov't budget is run so that the bonds are interest-only.
In general, household defaults mean the lender takes some kind of financial loss. If the gov't borrows in its own currency, it is expected that the currency printing entity would bail out the lenders (make the bonds whole).
Posted by: Too Much Fed | November 05, 2017 at 03:34 PM
I have to say that I still don't get. Why government debt is a burden but e.g. corporate is not? Maybe the answer is obvious like Tel said but I just don't see it (wouldn't be the first time)? Nick would you have time to also for this one? Or anyone else who gets it?
On other subjects:
"But keep in mind that it is easy to create financial assets, if their net value is not zero it would be easy to create infinite amount of wealth."
Nick and TMF, the backdrop here was that there is wealth creation by "refuse to accept" or "walking away". If that is the case, we truly can create an infinite amount of wealth.
Note that from financial asset > liability it follows that economic "total wealth" > "total real wealth". Is this a useful starting point? And also if assets == liabilities it doesn't follow that money couldn't have liquidity value.
And Nick, coming back to the old post, you started it by saying:
"Take Bitcoin for example. It's a financial asset to whoever holds it."
Why do you say Bitcoin is a financial asset? Wikipedia: "A financial asset is a non-physical asset whose value is derived from a contractual claim, such as bank deposits, bonds, and stocks.". Is Bitcoin "a contractual claim"?
Posted by: Jussi | November 06, 2017 at 08:05 AM
Jussi, how do you define wealth creation?
Also, let's try this scenario.
The gov't issues some 200 year bonds. The central bank buys the gov't debt. A corporation issues some 200 year bonds. Some, but not all, of the people buy the corporate bonds. The corporation also has stock. The corporate bonds are interest-only for 140 years. The gov't runs its budget/bonds as interest-only for 140 years. At year 100, everyone dies and there is a new generation. The new generation inherits the corporate stock. The new generation inherits the corporate bonds. At year 141, the gov't and corporation start making principal payments.
Where is the burden?
Posted by: Too Much Fed | November 07, 2017 at 02:15 PM
I have tried to summarize below the discussion on this matter:
Nick is saying that government debt is a burden, some economists (e.g. Krugman) are saying it is also an asset, meaning it is a wash - therefore the debt level itself doesn't matter. I think that Nick is right given his assumptions because if it is sold and not bequeath (in any form!) it has to be consumed by "the parents". But that is a big assumption because it means that more debt always means more consumption!
So, IMO, this is not "a biggie" or anything new - if debt is used to increase current consumption, it is a burden on future generations. Either the parents eat the apples or they bequeath them. So Nick's argument collapses back to the fact that it only matters whether the additional debt means higher consumption. And this applies to any debt class (e.g. corporate bonds). That is what Brad DeLong also said (2012 - answer to Nick): "What causes the burden is not that government debt is issued, but rather that the issuance of government debt crowds out the formation of useful capital." (http://delong.typepad.com/sdj/2012/10/the-intergenerational-burden-of-the-debt-nick-rowe-tempts-fate-weblogging.html).
Nick has said that Simon Wren-Lewis agrees with him. E.g. SWL: "Government debt can be used to redistribute income to current generations from future generations, even if the aggregate level of consumption in each period remains the same.". (https://mainlymacro.blogspot.fi/2015/02/the-burden-of-government-debt-again.html). The latter part, "even if ", should be highlighted as I'm not sure his comment is actually compatible with Nick's assumption that the debt is not bequeath in any way (meaning higher consumption). In other words SWL assumes here the kids are not worse off as their consumption is unaltered. So this is not a biggie. Sure enough he promptly continues: "The size of government debt is not a good indicator of any burden.".
Also Noah wrote that he and Nick "agree on everything now" (after disagreement): "These past government transfers can be accomplished by government borrowing and spending...This is the upshot of Nick's model.". Yet, Antonie Fatas (2012) seems to think that this is not a biggie: "My reading of the debate is summarized well by Noah Smith long list of updates to his blog entry. In particular the following question: is government debt an indicator of the (fiscal) burden we are imposing on the next generations? And the answer is a clear no. Debt does not matter. What matters is taxes and spending, debt is just a vehicle to deal with imbalances between the two. Debt is not a burden per se but it can be the outcome of tax and spending decisions that lead to redistribution of resources.". Read them yourself but I _think_ both Noah and Fatas think that this is not a biggie.
Nick (2015) himself later agrees with Fatas: "(Antonio Fatas, by the way, is correct in saying that if the government issues debt to buy real assets, and if future generations inherit (as a freebie) those real assets as well as the tax liabilities inherent in the debt, it can all cancel out.)" http://worthwhile.typepad.com/worthwhile_canadian_initi/2015/02/debt-does-have-intergenerational-distributional-implications.html. I think that "buy real assets" covers all public investments (e.g. roads).
Posted by: Jussi | November 08, 2017 at 10:49 AM