Let me try it this way.
Here's Brad DeLong:
"So, by continuity, somewhere between policies of austerity that that produce deflationary depression due to an excess demand for safe assets and policies of fiscal license that produce inflationary boom caused by an excess supply of government debt, there must be a sweet spot: enough new issues of government debt to eliminate the excess demand for safe assets and so cure the depression, but not so much in the way of new issues of government debt to produce destructive inflation, right?"
Is continuity assured?
(I'm sorry to pick on Brad, but he made the mistake of writing too clearly, and laying out the implicit assumption too explicitly.)
Then all of a sudden there was a discontinuous change in the market for private debt. Perceived risk jumped up. Demand jumped down. Prices jumped down. And those discontinuous changes caused a financial crisis, and that financial crisis caused the recession.
It wouldn't have happened under continuity. The prices of private debt would have fallen slowly as the stock of debt increased, causing the issuance of private debt to slowly grind to a halt, with no financial crisis.
Why couldn't the same discontinuous change happen in the market for government debt?
Yes, there is one big difference. Governments that own a printing press can and very probably will act as lenders of last resort to themselves. And they might not act as lenders of last resort to the private sector. That means government debt is less vulnerable to a liquidity crisis than private debt. But there are solvency crises as well as liquidity crises.
If people suddenly want to hold less government debt and hold more government money instead, the central bank can simply swap money for bonds and fix the liquidity crisis. But if people suddenly want to hold less government debt and hold less government money, that won't work.
Why can't there be discontinuity in the market for government debt too? Why can't there be a Minsky Moment with government debt too?
Yes, the results would be different: an inflationary crisis rather than a deflationary crisis. But it's still a crisis.
You want a model? Sure. Give me your model of the discontinuity in the market for private debt, I will replace "private" with "public", and hand you back the same model.
Update: and if you say that an economy at the ZLB is qualitatively different from an economy not at the ZLB, and many people do say that, you are saying that there is a discontinuity on the boundary between the ZLB and the non-ZLB equilibrium. "Too little, still a little too little, very nearly enough, far too much!" Back to (Ryan Avents?) Ketchup Theory. Monetary policy is easily and quickly reversible in a way that fiscal policy is not. The central bank can easily and quickly buy back the stock of money it has issued; the government cannot so easily and quickly buy back the stock of debt it has issued.
Dont you need the same level of uncertainty about public and private debt to make these two equivalent? I can see people not wanting to hold some kinds of private debt at any yield as the risks are perceived as too high. This seems much less likely with public debt, though I guess current circumstances could change that.
Steve
Posted by: steve | October 07, 2013 at 09:10 AM
Nick says, “Why can't there be discontinuity in the market for government debt too?” My answer: possibly there could be, but there aren’t any historical examples, are there? And the reason is that it would require a large and sudden shift in the types of assets that people want to hold.
You could cite Mugabwe and Weimar, but even there there wasn’t a SUDDEN attempt to escape money and get into real assets.
In contrast to government debt, private banks are INHERENTLY FRAGILE: they are far more heavily indebted than any normal corporation. That means that as soon as a rumour starts that a private bank is suspect, it is liable to collapse. Or as was the case with the recent crisis, when the entire shadow bank industry is suspect, there is run on the shadow bank industry.
I think I've said more or less the same as Steve just above, haven't I?
Posted by: Ralph Musgrave | October 07, 2013 at 09:22 AM
Steve: there is less risk of outright default with government debt (if the government owns a central bank), but the inflation risk is there. And history tells us that inflations do happen. Less risky than private debt? Yes, usually, especially since inflation will also reduce the real value of private debt, unless it's indexed to something. So a Minsky Moment in public debt will probably bring down most private debt too. But it's still uncertain.
Posted by: Nick Rowe | October 07, 2013 at 09:23 AM
"But if people suddenly want to hold less government debt and hold less government money, that won't work."
If both happen, the government can raise taxes and retire its debt until monetary savings and money are in equilibrium again with GDP and inflation.
Posted by: Odie | October 07, 2013 at 09:30 AM
Odie: yep. And how big a tax increase will be needed, if the government needs to retire a stock of debt suddenly? Will it be able to raise taxes that much that quickly? What are the costs if it does so?
Posted by: Nick Rowe | October 07, 2013 at 09:34 AM
"So, by continuity, somewhere between policies of austerity that that produce deflationary depression due to an excess demand for safe assets and policies of fiscal license that produce inflationary boom caused by an excess supply of government debt, there must be a sweet spot: enough new issues of government debt to eliminate the excess demand for safe assets and so cure the depression, but not so much in the way of new issues of government debt to produce destructive inflation, right?"
Focusing on the wrong thing. Excess demand for safe assets is because confidence and expectations are too negative becuase people dont have enough real income. Not becuase not enough treasuries are out there. If people had more real income their confidence and expectation would become more positive and lift prices of less safe assets. Focusing on government debt levels is wrong.
Posted by: Mike | October 07, 2013 at 10:30 AM
Why suddenly? Right now the assumption is that the size of the government debt in itself will be a cause for economic instability. However, if the government would monitor economic parameters like the Fed does (and include housing prices in its inflation stats!) it could adjust the size of its deficit/surplus accordingly. Maybe politically not the easiest one to implement but from an economics perspective a federal sales tax with a floating rate (e. g. adjusted every 3 months) would be a nearly perfect instrument to keep inflation under control due to the deficit becoming too high.
The question that I have now: Are economists concerned about the size of the government debt from an economic perspective or a political? Isn't that like saying:" Yes, evolution is right but we could never sell that to the public."? ("so we put some creationism in here and there to make it more amenable to political views").
Posted by: Odie | October 07, 2013 at 10:41 AM
Nick,
Just saw your update:"The central bank can easily and quickly buy back the stock of money it has issued.."
Are you sure it can? The Fed can "buy back" reserves. If there are no excess reserves, any further purchase would result in loans being called back by the banks. Would that be more economic stable than the federal government tightening fiscal policy?
Posted by: Odie | October 07, 2013 at 11:49 AM
As I understand it, the discontinuity in the private debt is due to the dual Diamond-Dybvig equilibrium (i.e. business as usual vs. bank run). I don't think any such dual-equilibrium model applies to government debt under fiat money. As someone already mentioned, even the worst hyperinflations occur as a gradual process, not a sudden phase shift.
You hint at this when you say that government can't get into a liquidity crisis. The answer to your question is that the "liquidity crisis" -- one of those much abused and endlessly confusing technical terms -- is precisely the way in which the discontinuity emerges in private debt markets.
Posted by: Vladimir | October 07, 2013 at 11:55 AM
Nick, the "liquidity vs. solvency" really may be the key here. There clearly are discontinuities with bank liquidity, but it's not clear there are any with government solvency. In particular, the actual revenue base for the government debt should be strongly procyclical, which gives us negative feedback on this process. Falling private demand for government debt (and the ensuing increased interest payments) is a positive feedback, but may not be strong enough.
Posted by: Alex Godofsky | October 07, 2013 at 12:07 PM
Nick,
I'm not sure that this really has anything to do with discontinuity. You don't need discontinuity to get cycles. When you have Ponzi borrowers, you can get a collapse just from having the demand for new debt peter out. The Minsky Moment might be actually be rather drawn out, but isn't that actually more realistic?
But otherwise, I'd agree that a sudden flight from public assets is likely to be inflationary.
Posted by: Nick Edmonds | October 07, 2013 at 12:21 PM
If the 2008 financial crisis was purely a liquidity crisis, and not a solvency crisis, then central banks caused the 2008 financial crisis by failing to act as lender of last resort, and failing to rescue solvent but illiquid financial institutions. There was no asset price bubble. OK. That's a defensible position. It may well be right. But I'm not sure how many people are saying that.
Odie: "Are you sure it can?"
Yes, I am sure it can (as long as central banks are solvent). If people (and banks) want to hold less government money, and more government bonds, so that failing to buy it back would cause inflation, the central banks can buy it back.
Posted by: Nick Rowe | October 07, 2013 at 01:23 PM
Nick, are there really multiple equilibria for solvency crises? Doesn't that imply that in some of the futures there wouldn't have been a crash? My understanding is that the same people you're talking about believed a crash was inevitable, which means that the non-crash state isn't really an equilibrium.
When you describe it as one, do you really mean that it's sort of a meta-stable state, where it can persist in the non-equilibrium for a reasonable amount of time before collapsing, and that sovereign date may be subject to the same dynamic?
(I'm just trying to get a clear idea of the thesis.)
Posted by: Alex Godofsky | October 07, 2013 at 02:16 PM
Nick Edmonds: Yep. If the demand for government debt declined slowly over time, it wouldn't be a big problem. The government could raise taxes a bit, cut spending a bit, and slowly reduce the level of debt.
Alex: the way I read the standard narrative, the argument is that the demand for private bonds persisted irrationally high for some time, then reverted to rationality. The bubble burst, in other words. Just one equilibrium, but the economy was away from equilibrium for some time. Personally I would prefer a multiple equilibrium explanation, and I would put more weight on monetary policy being able to keep us in the good equilibrium, if done properly, but I'm perhaps in a minority. But this post (unlike my last post) could be read either way.
Posted by: Nick Rowe | October 07, 2013 at 02:41 PM
Nick,
"If people (and banks) want to hold less government money, and more government bonds, so that failing to buy it back would cause inflation, the central banks can buy it back."
If people want to hold less money and more govt bonds, and if the central bank fails to buy back the money, why would people with the money suddenly want more goods and services (instead of govt bonds)?
I can see that they might want more private assets, and so buy those assets if the central bank fails to sell govt bonds, but why would this buying of assets necessarily lead to CPI inflation?
Posted by: Philippe | October 07, 2013 at 02:58 PM
Nick,
"If people suddenly want to hold less government debt and hold more government money instead, the central bank can simply swap money for bonds and fix the liquidity crisis. But if people suddenly want to hold less government debt and hold less government money, that won't work."
"Why can't there be discontinuity in the market for government debt too? Why can't there be a Minsky Moment with government debt too?"
Not sure about Canada, but the U. S. has a group of banks called the primary dealers that are required to submit bids on U. S. government debt issuance. See:
http://en.wikipedia.org/wiki/Primary_dealers
As long as it is profitable for a bank to do so - short term cost of funds from Federal Reserve is lower than long term interest rate on government debt, a bank essentially makes a risk free profit.
And so, if people want to hold less government debt and less government money, primary dealers buy up excess government debt, people dispose of government money by buying goods or reducing private debt.
"The central bank can easily and quickly buy back the stock of money it has issued; the government cannot so easily and quickly buy back the stock of debt it has issued."
Correct, but the government can roll over its existing debt on more favorable terms where the private sector may not be able to do so.
Posted by: Frank Restly | October 07, 2013 at 03:32 PM
But isn't inflation risk already factored in as price risk? We know that inflation can go up and government bond prices go down; we only need to look at the 1970s data. But once Bretton Woods was exitted (which some people view as a form of default, but which is not an issue at present), I think that there was no concern about getting paid back, just that the dollars you got paid back were going to be effectively worthless in real terms. Debt to GDP ratios were low during that period, due to the high nominal GDP growth rate. Based on that experience, if I am worried about inflation, I do not simultaneously worry about the fiscal ratios.
I see this as different from the problem faced by private borrowers in 2008, whose problem was that they lacked safe, liquid (government backed) assets on their balance sheets.
Posted by: Brian Romanchuk | October 07, 2013 at 04:03 PM
Philippe:
OK, let's say all of the people who want govt bonds settle for MSFT stock (take as proxy for private assets in general) instead:
In this case, they can't actually increase their stock of MSFT, they can only bid up the price. Let's assume for the moment that expectations about future MSFT revenue/profits/dividends/etc. do not actually change (why would they? MSFT isn't doing anything different than it was going to before). Then the holders of MSFT would know that they are getting fewer dollars of future consumption per dollar of MSFT than before. At some point this should cause some people to spend their marginal dollar on present consumption instead of future consumption. More money spent on present output => higher price level!
This is identical to people just holding onto the dollars, which per assumption they don't want to do (they want to hold onto less money).
More money spent on present output => higher price level!
Posted by: Alex Godofsky | October 07, 2013 at 04:04 PM
"It wouldn't have happened under continuity. The prices of private debt would have fallen slowly..."
The fact that a function is continuous does not imply that it varies slowly, and vice versa.
The continuity argument seems bogus to me: you can use it to prove a theorem, but here we are speaking of possible actions of a government or a central bank: if the economic system state varies quickly enough the 'sweet spots' might exist in theory but be undetectable / unusable in practice.
Posted by: MiMo | October 07, 2013 at 04:35 PM
In the private case, bad underwriting left the value of existing debt questionable and of much less value in fact, but I do think these same sudden changes in expectations can occur in the public case. The bank will then have to buy some debt and let rates rise on the rest which causes a fiscal crisis leading to spending cuts and increased taxes that will lead to a primary surplus. The debt won't ever be liquidated at once, only shown that it will grow less than the economy, but the economy should be doing well so it shouldn't present an overwhelming problem. When the economy has been hit with a negative shock and isn't doing well, generating less revenue and more stabilizer spending for the government, public debt will be seen as safer. Economic crises are slope discontinuities rather than level discontinuities. Change may be painful, but not catastrophic.
Posted by: Lord | October 07, 2013 at 05:01 PM
Nick,
Not sure I understand your "update", but this bit seems wrong:
"Monetary policy is easily and quickly reversible in a way that fiscal policy is not"
Say the Fed faces 3.5% inflation expectations and UE stuck at 7% while it is buying $100b in Treasuries a month. Is monetary policy "reversible" in this situation? Theoretically, yes; practically, no.
Posted by: Anon1 | October 07, 2013 at 05:31 PM
Nick,
Now I am getting confused. "Yes, I am sure it can (as long as central banks are solvent). If people (and banks) want to hold less government money, and more government bonds, so that failing to buy it back would cause inflation, the central banks can buy it back."
If the Fed can sell its government bonds does that not mean the public has a demand for them? Would that not be the normal equilibrium with the Fed buying and selling Tsys depending on the demand for money by the public? I thought the problem would occur when the public neither wants to hold money nor bonds. (With other words does not want to hold monetary savings but rather real assets instead.)*
In addition, the Fed usually (e. g. before 2008) holds only a small fraction of government bonds. If the demand of the public for money goes down (savings go up) it can sell only as much of them as it has as assets on its balance sheet. It should be close to M0 minus cash. If it would do that banks would only buy as much bonds as they have excess reserves or they would be non-compliant with regulations. If the Fed wants to force them to buy more bonds (or increase the reserve ratio) banks would need to decrease their deposits by calling back loans in order to stay within their reserve requirements. I believe that would be a severe shock to the economy probably as bad or even worse than the government increasing taxes/reducing spending.
* I can see a scenario where that would be happening and that would be during a severe supply crisis when prices rise suddenly because real output of goods and services breaks down (your sunspot?). However, I would consider this an external event and not something high public debt precipitated (although it may make it worse). In addition, I doubt that KR is worried about that eventuality; neoclassical models usually assume such events away (sorry, a little bit of sarcasm). If monetary stability is the main goal then the government could still run a surplus to curb inflation. However, it will not change the fact that there will be less to consume overall and most people will be worse off. (As far as I know, a good number (most?) of hyperinflations were actually precipitated by supply shocks to which the government responded by running the presses instead of fiscal tightening making a bad situation even worse.)
Posted by: Odie | October 07, 2013 at 09:28 PM
Nick,
"If the 2008 financial crisis was purely a liquidity crisis, and not a solvency crisis, then central banks caused the 2008 financial crisis by failing to act as lender of last resort, and failing to rescue solvent but illiquid financial institutions. There was no asset price bubble. OK. That's a defensible position. It may well be right. But I'm not sure how many people are saying that."
I think the 2008 crisis was more than just a liquidity crisis. Loans on mortgages went into default and interest rates also increased steadily. Both did lead to a reduction of bank capital to the point where banks became insolvent. Why did consumers default on their mortgages? I may be pretty alone in my position but I think the huge trade deficit played a major role. In the years leading up to the crisis the US exported between $500 to $700 billion per year. That money was missing in domestic income and could only be made up with taking on more debt. Those bonds were provided by the savings of our foreign trading partners thereby keeping interest rates of CDOs (together with the Fed) artificially low. At some point that debt level became unsustainable since the income to service those debts was just not there, investors pulled out from CDOs and the rest is history.
The Fed could maybe have stopped that by buying CDOs in large enough quantities but that is a political decision, not something the Fed should have done by itself.
Posted by: Odie | October 07, 2013 at 10:04 PM
Alex,
I more-or-less agree with your case 2, however, I don't think that a decision by businesses to increase investment would necessarily result in inflation if there is a large output gap and high unemployment. Also, i don't think that businesses will necessarily decide to increase investment because there is a higher demand for their shares or bonds. They will increase investment, it seems to me, if they think that there will be a greater demand for their goods and services from the public in the future.
Your case 1 sounds a lot like the "wealth effect" touted by the Fed. I don't think that this would have much of an effect on AD if the assets we are talking about are stocks and shares, given that not many people (as a percentage of the total population) change their spending patterns on the basis of changes in share prices.
Posted by: Philippe | October 07, 2013 at 10:08 PM
Philippe:
That's as may be, but it's a far cry from what you were asking about. Yes, under some circumstances monetary stimulus will mostly increase output and not do much to inflation, but the transmission mechanism to inflation is pretty clear.
They have to per the premise of your question. Remember, this is what you asked:
If the corporations just hold onto the money, your premise fails. You have to assert that people continue to prefer money to any kind of real investment whatsoever.
Posted by: Alex Godofsky | October 07, 2013 at 11:49 PM
"the transmission mechanism to inflation is pretty clear"
Not really. You didn't specify why companies choose to increase investment, you just suggested that they might, for some reason. I don't think they necessarily would as the result of "monetary stimulus".
Posted by: Philippe | October 08, 2013 at 12:05 AM
just to be clear, when I said that I agree with your case 2, I meant I agree that increased investment would probably have some effect on the price level, but I'm not sure it would necessarily be significant if there is high unemployment/ a significant output gap. I was disagreeing with your assertion that increased investment necessarily = inflation.
Posted by: Philippe | October 08, 2013 at 12:11 AM
"If the corporations just hold onto the money, your premise fails. You have to assert that people continue to prefer money to any kind of real investment whatsoever."
Again, not necessarily. The result could simply be higher asset prices (assuming people buy private assets in response to a 'lack' of govt bonds). Your case 1 asserts that some sort of 'wealth effect' results in more spending on goods and services as the result of higher asset prices. I agree that this is a possibility, but I don't think it is likely to be significant if the higher asset prices are predominantly realised in stocks and shares. If there is a large general increase in house prices, then I would agree that the wealth effect could be quite large. But such an effect would probably be quite short lived, as it is not based on an increase in sustainable income.
Posted by: Philippe | October 08, 2013 at 12:38 AM
@ Alex Godofsky,
There's a third case. The average price of investment vehicles could increase. If you assume neutrality of money in a market containing only investment vehicles, then you can see this is possible.
What determines the relative inflation rates of the CPI and financial assets is an important and interesting question for which I'd love to see an answer.
Posted by: PeterN | October 08, 2013 at 04:00 AM
Nick,
Have you considered mark to market rules as a factor? They convert the stock of unrealized losses into a flow of realized losses. These rules were an attempt to prevent hiding unrealized losses by putting them in accounts for assets not intended for sale.
This is obviously a destabilizing positive feedback. These accounting losses can force the sale of assets, driving down prices and causing further mark to market losses. We saw this in the sub-prime market. Regardless of any fundamental value, investors won't take the risk of buying into a rapidly falling market which "can remain irrational longer than you can remain solvent".
Posted by: PeterN | October 08, 2013 at 04:13 AM
Philippe: because the LM curve shifts right, from what was initially (by assumption) a level of output at which inflation was on target. Or, what Alex said.
Brian: you could also argue that default risk should have been factored into private bond prices in 2007. The standard narrative says it wasn't.
MiMo: OK, a discontinuity ( a large jump down in the demand for government bonds) would be sufficient for my argument. A continuous but rapidly declining over time demand would also have much the same effects.
Lord: maybe, I don't know. But I think there's too much complacency. And it's rather ironic given what happened to the demand for private debt.
Anon1: in that case there might be a question of whether you want to reduce AD or not. But nevertheless, if you did want to reduce AD, a solvent central bank can immediately buy back all its outstanding stock of money, if it wants to.
Odie: you lost me. Let me repeat: a solvent central bank has assets (mostly bonds) greater than its monetary liabilities. That means if it wants to buy back all its monetary liabilities it can do so.
PeterN: maybe. But I don't think you can look at mark to market rules, without looking at rules on capital ratios plus banks' expectations of how those rules will interact.
Posted by: Nick Rowe | October 08, 2013 at 05:32 AM
Philippe:
I don't have to. If MSFT issues new shares it can either put the money in a box or buy something with the money. Per your assumption they do not put the money in a box. ergo, they buy something. QED.
If they do put the money in a box then we just keep printing money until people no longer put it in boxes.
Posted by: Alex Godofsky | October 08, 2013 at 09:34 AM
Philippe:
No, case 1 does notrelative price of present consumption and future consumption shifts in favor of present consumption, then some people will choose not to purchase future consumption and instead purchase present consumption.
PeterN:
Dude, that's my Case 1. And no, there isn't a third case, the two cases are "MSFT issues new shares" and "MSFT doesn't issue new shares". What's your third case, aliens invade and buy all the shares to use as toilet paper?
Posted by: Alex Godofsky | October 08, 2013 at 09:38 AM
Sorry, something got mucked up with my html tags. Above should read "No, case 1 does not assert a wealth effect. If the relative price".
Posted by: Alex Godofsky | October 08, 2013 at 10:14 AM
Nick,
You say:"Anon1: But nevertheless, if you did want to reduce AD, a solvent central bank can immediately buy back all its outstanding stock of money, if it wants to.
Odie: you lost me. Let me repeat: a solvent central bank has assets (mostly bonds) greater than its monetary liabilities. That means if it wants to buy back all its monetary liabilities it can do so."
The Federal Reserve has currently ~$3.7 trillion in assets on its balance sheet which is unusually high. (http://www.federalreserve.gov/releases/h41/current/h41.htm#h41tab11) Of that, ~$2 trillion are government bonds. Before the 2008 crisis, those figures were in the ~$800 billion range (http://www.federalreserve.gov/monetarypolicy/bst_recenttrends.htm). Since the Fed can only sell as many bonds (buy back its money) as it has assets on its balance sheet, the FED could never buy back all the outstanding stock of money with M2 in the $10 trillion range. The treasury can always sell government bonds for money; the FED only in the amount it bought previously through its OMOs. (Btw. The Bank of Canada has only treasuries for 87 billion dollar on its balance sheet; would that be enough to buy back all the outstanding stock of money?)
But again, we are not really talking about your discontinuity/Minsky moment:"But if people suddenly want to hold less government debt and hold less government money, that won't work." Here people don't want neither and the FED cannot sell bonds to reduce its outstanding money stock because people do not want to hold bonds. So, how do you get rid off both? By taxing people out of their money and using the surplus to reduce the outstanding public debt (=supply of government bonds). IMHO, that will be the only available strategy in the Minsky moment you described but quite frankly not so difficult if people would actually understand that mechanism.
Posted by: Odie | October 08, 2013 at 12:59 PM
Odie: M2 is not a liability of the central bank.
Your comments are beginning to annoy me. They are taking up space, and taking up my time, without really adding to the conversation.
Posted by: Nick Rowe | October 08, 2013 at 01:03 PM
Nick,
That's certainly true. But when the Fed cannot buy back all money it means there could still be inflation when people do not want to hold money nor bonds anymore. Looking over the posts of some other commenters I am certainly not the only one who has troubles with your assertion that the Fed can more easily control the discontinuity you are talking about. But if you don't want to discuss that fine, it's your blog.
However, I also took an important (bolded) statement from you and tried to argue that monetary policy will not work in that instance only fiscal policy can. Instead of discussing the issue you are now attacking me personally. I thought open discussions were part of the academic discourse to improve our knowledge. That is at least how I deal with my students and colleagues.
Posted by: Odie | October 08, 2013 at 02:13 PM
@Alex Godofsky,
The difference is in the second half of your case 1, where you say:
" Then the holders of MSFT would know that they are getting fewer dollars of future consumption per dollar of MSFT than before. At some point this should cause some people to spend their marginal dollar on present consumption instead of future consumption. More money spent on present output => higher price level!"
If there were some formula for this, then financial asset prices and the CPI would move in lockstep. They don't. It seems that people haven't read the Econ 101 stuff about marginal prices. Certainly at some point some portion of people will probably spend some additional money on present consumption ceteris paribus. However we have no good way of predicting what this portion or point might be, and all things otherwise are rarely equal.
If people see the price of shares rise, they may feel that the discounted future utility of their estimate of the ROI exceeds the utility of current consumption. We know, after all, that the average family's investments are far short of what is needed to maintain their current lifestyle (or anything like it) in retirement. We've also seen a decline in the replacement market for household durable goods, even though their real prices have fallen dramatically over the last 40 years.
An additional dollar added to the economy will distribute its effect between RGDP, CPI inflation and asset inflation, but we don't seem to have a good way to know exactly what this distribution will be.
You can just as easily make the argument that people need a certain return from their investments, and if the yield falls, they will invest more to maintain it. You have to consider the relative elasticities of current and future consumption and peoples' expectations of future earnings (which seem to be declining). Between that and people buying on expectations that current rises in asset prices imply similar future rises, it's hard to make predictions. If it weren't then the observed volatility would imply extremely inefficient markets.
Posted by: PeterN | October 08, 2013 at 05:48 PM
@Alex Godofsky,
"I don't have to. If MSFT issues new shares it can either put the money in a box or buy something with the money. Per your assumption they do not put the money in a box. ergo, they buy something. QED.
If they do put the money in a box then we just keep printing money until people no longer put it in boxes."
S&P 500 companies as a group aren't selling new shares, they're buying back (100s of billions of dollars of) existing ones and borrowing money to do it - inflating the price of their stock with borrowed money.
If MSFT does buy something, it will be another company, for which they will pay a premium.
Posted by: PeterN | October 08, 2013 at 05:59 PM
PeterN:
No, because relative prices are changing all the time. But hey, guess what, equity markets and inflation expectations have been positively correlated for the past few years. Whoda thunk?
Aside: this is rich, coming from you.
I could apply your statement to literally every single claim made in any micro- or macroenomics textook ever and it would be equally true and useless.
Nope! Wrong! This is what happens when you don't think rigorously about how price changes tell us about shifts in supply and demand curves.
Remember, in Case 1 land the TOTAL QUANTITY of the investment good (measured in future consumption) was unchanged. All that happened was that its PRICE increased. That means that everyone who was already holding the investment good still has exactly the same expected future consumption from it.
So your thesis is that a rightward shift in the demand curve for future consumption reduces expectations of future earnings... why?
Yes, if you simply assume that the entire market is inefficient, you can conclude that the market is inefficient!
"MSFT" was explicit placeholder for "private financial assets of all stripes". It includes corporate debt etc. If they buy another (existing) company that doesn't actually answer the question of what happens to the money - if the money ends up being put in a box then we've just rejected the original premise: "If people want to hold less money and more govt bonds".
Posted by: Alex Godofsky | October 08, 2013 at 08:33 PM
"Remember, in Case 1 land the TOTAL QUANTITY of the investment good (measured in future consumption) was unchanged. All that happened was that its PRICE increased. That means that everyone who was already holding the investment good still has exactly the same expected future consumption from it."
How does someone determine the "TOTAL QUANTITY of the investment good (measured in future consumption)". Assuming the existence of such a thing as an available factor for decision making is assuming what you seek to prove.
"No, because relative prices are changing all the time. But hey, guess what, equity markets and inflation expectations have been positively correlated for the past few years."
Yes, but the usual correlation is negative. I'm not that interested in the last few years as a special case. I'm interested in determinants of the relationship between RGDP, CPI inflation and asset prices in general.
It is obviously true that higher asset prices induce higher goods prices to some extent eventually, if only through a wealth effect, but the huge differences between the rates of increase at certain times could use an explanation.
". If they buy another (existing) company that doesn't actually answer the question of what happens to the money - if the money ends up being put in a box then we've just rejected the original premise: "If people want to hold less money and more govt bonds"."
Aren't you mixing stocks and flows?
Posted by: Peter N | October 08, 2013 at 11:44 PM
"Monetary policy is easily and quickly reversible in a way that fiscal policy is not."
I don't think this is true. (Indeed, I think the whole advantage of fiscal policy is that it can be reversed in situations where monetary policy cannot.)
There is no guarantee that there is an equilibrium consistent with the monetary policy target (whatever that target is, but in the most relevant case, if the CB is targeting 2% inflation, the natural real interest rate may be less than negative 2%). Therefore in order to achieve its target, the CB may need to implicitly change its target. That is, it might need to credibly commit to a different target. Is such a credible commitment reversible? If it were reversible, one would have to wonder how it managed to be credible in the first place.
As for fiscal policy, I suppose the question would be, if you get out of a depression and then decide the alternative is worse, can you get back into the depression? If your debt is denominated in your own currency, I don't see why not. You can tighten fiscal policy enough that there is no longer inflationary pressure. Then next time you don't bang quite as hard on the ketchup bottle.
Posted by: Andy Harless | October 09, 2013 at 02:48 PM
Andy: some targets are infeasible, because they are not equilibria. Sure. We restrict ourselves to targets that are feasible, if we are sensible.
It is one thing to change the flow of new bonds issued quickly. It is quite another thing to change the stock of outstanding bonds quickly. The central bank can change the mix of stock of bonds and money almost instantly.
Posted by: Nick Rowe | October 09, 2013 at 03:22 PM
It is quite another thing to change the stock of outstanding bonds quickly. The central bank can change the mix of stock of bonds and money almost instantly.
The real problem is that to dramatically tighten FP enough to whip the economy back into the depressed equilibrium the government needs to do things that are politically impossible (even if there is a world in which they would not in fact violate the long term budget constraint if passed and could theoretically issue a large lump sum tax in good time), and the market will know how hard this would be. This is why I mentioned in the previous thread that a potential solution to this is to use fiscal policy in the depressed equilibrium to borrow to accumulate a portfolio of real assets. These assets can, and the market would realize they can, be very quickly unwound to repurchase and dramatically reduce the stock of government debt in the good equilibrium just as the CB can unwind its portfolio very quickly and reduce the stock of base money as it tightens.
You can use FTPL type language to say this initially works by increasing the riskiness of government debt (rather than framing it as meeting the high equilibrium demand for debt) at the ZLB but allows for a policy reversal that is contingent upon a good equilibrium. This is just the government taking the other side of the twin equilibria risk premia trade in two states of the world (presumably a profitable externality closing trade).
Posted by: dlr | October 09, 2013 at 07:25 PM
Nick,
"And how big a tax increase will be needed, if the government needs to retire a stock of debt suddenly? Will it be able to raise taxes that much that quickly? What are the costs if it does so?"
A government could also sell equity and use the proceeds to retire debt.
Posted by: Frank Restly | October 09, 2013 at 11:20 PM
In theory there is a difference between discontinuity and rapid change, in practice there may not be. ;)
In any event, Brad DeLong is right about a continuity of **policies** (at least to very small differences). And Nick Rowe is right about a possible rapid change in mass psychology. Possibly such a change would necessitate a change in policy.
And if there is such a change, then we have to talk about the changing conditions that cause it, right?
Nick, are you envisioning a chaotic situation where a small change in policy could result in a large change in mass psychology?
Posted by: Min | October 09, 2013 at 11:22 PM
dlr: "This is why I mentioned in the previous thread that a potential solution to this is to use fiscal policy in the depressed equilibrium to borrow to accumulate a portfolio of real assets. These assets can, and the market would realize they can, be very quickly unwound to repurchase and dramatically reduce the stock of government debt in the good equilibrium just as the CB can unwind its portfolio very quickly and reduce the stock of base money as it tightens."
I rather like that idea. Or, we could have the central bank buy those real assets. And we could argue about where along the spectrum "real" assets begin, and where monetary policy ends and fiscal policy begins. But those semantic questions matter less than the economic effects.
Min: I think so. I think I see a sort of backward-S-shaped demand curve for government bonds, so a slow increase in supply would cause a sudden drop in price as the economy flipped out of the ZLB equilibrium into the normal equilibrium. Or a slow shift in the demand curve to the left would do the same thing.
Posted by: Nick Rowe | October 10, 2013 at 04:47 AM
Nick Rowe: "I think I see a sort of backward-S-shaped demand curve for government bonds, so a slow increase in supply would cause a sudden drop in price as the economy flipped out of the ZLB equilibrium into the normal equilibrium. Or a slow shift in the demand curve to the left would do the same thing."
Thanks. :) I haven't really studied it, but I think that such a demand curve gets us into catastrophe theory. :)
But, as Ralph Musgrave indicates, it's an empirical question, isn't it?
Posted by: Min | October 10, 2013 at 10:45 AM
> If people suddenly want to hold less government debt and hold more government money instead, the central bank can simply swap money for bonds and fix the liquidity crisis. But if people suddenly want to hold less government debt and hold less government money, that won't work.
> Why can't there be discontinuity in the market for government debt too? Why can't there be a Minsky Moment with government debt too?
I'm not sure that this is the right question to ask. If people suddenly want to hold both less debt and less money, then what do they want to hold instead? After all, money has value relative to other things.
One potential answer which comes up historically is foreign money. Through the black and grey markets, foreign money ends up capturing a large segment of a nation's economy, and then the sovereign currency is no longer the de facto national currency. The resulting hyperinflation would be the flip side of a Greek-style debt crisis. This is also why countries that deliberately devalue their way out of debt have to impose capital controls.
Another potential answer to the preference is "hard goods." This situation will arise if shortages are present or expected -- holding money becomes useless because you won't be able to get X in the future, for any price. That seems more the situation with the Weimar-era hyperinflation, as the trigger appears to have been reparation payments in gold or hard goods, which were then unavailable to the national economy. Since the government attempted to monetize its way out of those reparations (through market purchases of gold and foreign currency), money was no longer a stable store of value.
In both cases, however, I'm not sure the Minsky Moment is the critical thing. Instead, it would be more a symptom of an underlying disease.
In the particular and peculiar case of the US, I think that government panic over the debt could ironically effect the very crisis. The US economy is too large to be dominated by a foreign currency, but widespread belief in a crisis-facilitated depression could cause real-goods hoarding and shortages.
Posted by: Majromax | October 10, 2013 at 03:01 PM