Wednesday, January 30, 2013

Where is the money going to come from?

Megan represents responsible economic management, which means that she follows more-or-less textbook micro and macro and has the concerns common amongst DC-centric administrators. It's a mix of Chicago-style modeling with upstate New York pragmatism and the sweep of a DC policy wonk.

Recently she wrote a reasonable piece about why worrying about growth was not enough, you also need to worry about the deficit:
So why are people focusing on the tedious and painful business of austerity, when growth would be so much better?  For the same reason you've probably opted to pay off the Mastercard, rather than waiting until you have time to publish a bestselling novel: it's not so easy to deliver robust economic growth on demand.  Whatever you may have heard, no one has a plan in their pocket to increase the trend rate of economic growth--indeed, so far we've failed to get it back to the levels that preceded this "one time factor".  Telling budget wonks that "we need more growth" is a bit like telling a cancer patient "you need more health".  I mean, yes, Dr. Insight, but can you be more specific?
To be sure, we do know how to boost growth in the short run: borrow a bunch of money and throw it into the economy.  But this is exclusively a short term strategy.  Moreover, stimulus doesn't fix our budget problems; it increases them.  The current federal tax take is somewhere south of 20% of GDP.  That means that for stimulus to pay for itself, budget-wise, it needs to have a multiplier of 5--which is to say that every dollar the government spends must generate $5 worth of GDP growth.  Recent estimators of the multiplier during the Great Recession were more like 1.5, which means that for every dollar we spent on stimulus, we generated an additional 10 cents in tax revenue.  This is not a financing strategy that can be kept up forever.  A lot of liberals seem to be thinking of stimulus the way that some conservatives think of tax cuts: as a sort of perpetual motion free money machine.  There is no such thing.
This is all reasonable and responsible and something that people from both sides of the political aisle would broadly agree with. It is also, however, wrong. Currently, the output gap in the US is not coming because we aren't inventing new iPhones, it's coming because of a lack of aggregate demand. We have a lack of aggregate demand because the private sector's savings desire is not being met. The private sector's savings desire is not being met because the Government is not printing the money people want to put in their bank account, so they are trying to save from each other and, of course, as a sector cannot. The Government is not printing the money people want to put in their bank because it thinks it 1) cannot, 2) if it does, it will inevitably result in inflation, and 3) even if it does, it will need to pay it all down someday and the number is already so big.

None of this is true. Fiat currency is a marker of sovereign power, and a sovereign which chooses to tax needs to supply currency the same way that a sovereign which chooses to imprison needs to issue laws. A failure to print (and unprint) is a failure to rule, and is ultimately the irresponsible position to take.

Today's bad GDP number provoked another article, where Megan sees the aggregate demand problem but does not see the solution which sits between her two pieces.
The good news, such as it is, is that personal consumption spending and investment were humming along, growing 2.2% and 8.4%, respectively.  The boost in personal spending was driven mostly by durable goods, which probably means that people are reaching the limits of hoarding--they've pushed the old car along an extra five years, put up with the oven that doesn't always work, and gone without a dishwasher, but they're now having to replace some stuff.  
In theory, that can touch off recovery, as production eventually ramps back up, and rising confidence ripples through the economy.  (Paul Krugman had a great explanation of this a while back, but I can't find it.  Curse your prolific output, Professor Krugman!)  But in the context of an otherwise lackluster GDP report, this is worrying in the short term: people aren't buying because they have more confidence in their future, but because they feel they have absolutely no choice.  We don't want to be in a place where people reluctantly pry open the piggy bank only because the old clunker is finally lying smoking in the driveway; we want an economy where people feel that it is safe to buy a new car, because they are likely to have a job in three years.
Businesses hire when they have customers. Business have customers when the customers have money they are willing to spend. Customers have money they are willing to spend when their piggy banks are nice and full. Their piggy banks become nice and full when the Government prints money to put in their piggy bank. Cannot come from anywhere else.

(To the pedants out there, and you know who you are, note that I am speaking in the context of the US case, and by "money" I mean net financial assets (equity) denominated in US$, by "Government" I mean the US Government and am assuming that the Treasury and Fed can act in tandem with each other and that Congress will govern, and by "business" and "customers" I mean and non-US Govt sector which would include non-US businesses and individuals who wish to transact in US$ as their specie of choice).


Blogger ds said...

McArdle's logic is so convoluted it makes a buffoon such as myself feel smart.

You don't need a graduate degree in economics or a mastery of stocks and flows to figure out the logic here -- all you need is a little common sense. The Paul Krugmans of the world say "the fiscal multiplier is 1.5 -- more fiscal stimulus is warranted". The Robert Barros of the world say "the fiscal multiplier is zero, fiscal stimulus is not warranted". So if McArdle has it in her head that the fiscal multiplier needs to be FIVE in order for fiscal stimulus to pay off, then why in the world do all the Robert Barros of the world fight the Paul Krugmans tooth and nail over the belief that the multiplier is 0 and not 1.5?

5:12 PM  
Blogger Roman P. said...


Maybe it is better to say that the money the customers could spend might come either from the banks extending credit (in which case the Govt passively extends enough reserves to the banks) or from the government spending on goods and services?

Then, if banks are unwilling to extend credit and firms are unlikely to borrow, the government has to spend to compensate for the lack of demand. That'd be a broadly Post-Keynesian position; you MMT guys sometimes acknowledge it, but sometimes word it in such a way that suggests that all money comes from the government and the government alone decides what the money supply is. Why do you do that if that leaves you open to the critics calling you quasi-monetarists?

3:58 AM  
Blogger winterspeak said...

ds: "fiscal multiplier" is not a helpful way to think about the effect of stimulus.

roman: I don't think it's better to say that because it isn't accurate.

The non-Govt sector wants net financial assets, not gross financial assets. Bank lending generates gross assets, not net assets.

And the Govt does not have to spend, it can also tax less.

Bank lending is not reserve constrained.

9:12 AM  
Blogger paul meli said...

"the government has to spend to compensate for the lack of demand"

The government has to spend or there would be a lack of demand, not to compensate for it.

The demand businesses create is equal to their gross expenditure. This is not even close to enough funds to purchase their products so where else will the funds come from if they are to make a profit?

And don't say credit…this is like saying we should loan money to players in a poker game…the winner takes all…how would the other players pay it back?

10:40 AM  
Blogger Roman P. said...

1. If you could explain or provide a link to the explanation, perhaps? I've seen this idea in MMT writings, but its meaning remained obscure to me. From what I understand, banks create money by the acts of lending, and money is money and could be used to buy things. Everyone could credit the others and create money or money-like things (IOUs, private currencies, etc), actually. These acts are, of course, not directly constrained by the actions of government.
2. Let's not get into that for the moment.
3. Of course bank lending is not constrained by reserves, this kinda allows (1). You think I'm holding an opposite view?

12:12 PM  
Blogger winterspeak said...

Roman: When a bank makes a loan is expects to be paid back, so a loan extension generates a receivable (asset) plus a deposit (liability) in the same sector, netting out to zero.

So it creates gross financial assets (one kind of "money") but not net financial assets (a different kind of "money"). If I borrow $100 it's different than when I find $100 lying on the ground, even though in both cases I have $100 more in my wallet than I did before.

For the non-govt sector to increase it's amount of this kind of net money, it must come from out of sector.

1:53 PM  
Blogger Roman P. said...


Imagine that you borrow $100 from the bank and immediately withdraw the same sum from your account in cash. Your bank calls up the central bank and exchanges the money it created into the government money (banknotes). And gives them to you. Now you have $100 more in your wallet, and they are no different from the cash you could have found on streets, received by mail from the government or acquired by any different means. At the moment, they increase your ability to spend. You, of course, will have to repay $100+N, reducing your effective demand, but at the later date. My point is that injections (getting money from the banks) are not immediately offset by the leakages (repaying the debt), and this delay matters.

Have you seen Keen's little dynamic model of the economy with only firms, workers and banks? No public sector, so it only models a horizontal channel of money creation. One might think that the agents will go bankrupt by having to repay their debt with interest, but Keen's main result was that the model was stable. The money that the banks created by extending the credit generated a lot of wealth by circulating in the economy (a lot more than the size of the credit), only a small part of which went into the payments to the banking sector.

12:54 AM  
Blogger NeilW said...

To exchange the money a bank created the bank had to give something to the central bank. What did it give?

1:46 AM  
Blogger paul meli said...

"The money that the banks created by extending the credit generated a lot of wealth by circulating in the economy (a lot more than the size of the credit), only a small part of which went into the payments to the banking sector."

The money circulates but the level goes down as payments are made...requiring more debt, resulting in an unsustainable upward spiral. Saving reduces the level further. Saving is the fly in the ointment. Debt can't fund saving in the aggregate.

Small part? 10% of the outstanding loan balance annually is not small...that's roughly $1.3T a year from household debt alone. Half of that disappears into the ether reducing aggregate demand that must be replenished by the fiscal authority if the level is to be maintained.

Saving removes funds necessary to make payments. I won't go into the absurdity of borrowing the funds necessary to make the payments too. Borrowing from Peter to pay Paul. Even if that might work for a while it's insane behavior.

In the aggregate it may seem to work until the level of indebtedness gets high enough where aggregate debt service can't be maintained by aggregate income. That's the saturation point, which we have reached.

If one looks at the history of household debt as a percentage of $NFA a ceiling will become apparent...we have never broached that ceiling without suffering a banking crisis.

Business debt is extinguished through sales so households ultimately are responsible for all indebtedness. It all comes down to fiscal injections because households can't provide enough aggregate demand from wages earned through private business alone to clear inventory. Not.Even.Close.

The government must purchase a significant portion of output for capitalism to succeed.

There are reasons why we've run deficits more than 80% of the time since 1930.

5:56 AM  
Blogger Roman P. said...


I will be honest with you, I have no idea concerning US. From what I know of the Russian system, though, banks make an agreement with the central bank that allows it to exchange the rubles bank created 1:1 with the banknotes of the central bank within the certain limit. The flow of cash goes both ways: whenever the bank is paid by the customers in cash or when the firms return the cash from their cashier's offices, it converts that cash into the (virtual) rubles on their accounts by shipping it back to the (I'm probably butchering the real translation of the term here) CB's cash settlement centers.

I wanted to reply to your previous comment too, but decided against being rude then. Do you read my comments with your ass? It's as if you read the word 'debt' in my comment and immediately started to ramble on.

Apologies to the owner of the blog.

7:10 AM  
Blogger NeilW said...

Yes but the accounting entries are that the bank has to give up an *asset* to the issuing authority in exchange for the *asset* 'cash in vault'. And that asset is generally bank reserves - previously created by some sort of government sector spending and owed to the private bank by the central bank.

Once the private bank has cash in vault it can then write out the customers liability against that cash in vault asset reducing both. So the liability that was offset against the bank reserves is now offset against the created bank loan instead

The customer that was owed by the private bank is now owed by the central bank - and the private bank is no longer owed by the central bank.

The point being that the customer may have got cash, but that is because there was due to the existence of a previous NFA injection that created the bank reserve asset.

So in reality when a bank issues cash to a customer all they are doing is shuffling the existing liabilities and assets around. No new government sector injections occur.

7:58 AM  
Blogger winterspeak said...

Roman: Steve Keen is misleading on this point, and he does not distinguish between currency issuers and currency users. There is no separation in his model between vertical money created by an out-of-sector actor (Govt, which has a sovereign printing press) and horizontal money created by in-sector actors (banks).

Cash should be thought of as inventory. It has some interesting properties, but does not help in understanding the key point here.

If you borrow $100 from a bank, whether you then withdraw the money to put in your wallet, or just keep it in your checking account, or have it in your checking account for a little while until it gets transferred to someone elses checking account (maybe you buy a nice watch and you borrow from the bank to pay for it), you still need to pay that $100 back. Whatever asset this represents on your personal balance sheet--cash in wallet, checking account balance, or watch--it is balanced by a non-equity liability on your balance sheet as well which is the $100 you owe to the bank.

If you find $100 on the ground then it lives again as an asset of some sort--depending on what you do with it--but there is no corresponding non-equity liability on your balance sheet. Therefore, on the liability side, you book it as equity.

It is this equity entry on the liability side of the balance sheet I am talking about when I say Net Financial Assets (Equity). At a sectoral level, it cannot increase as assets must always balance liabilities -- anything created on the asset side must be balanced by something on the non-equity liability side. If it comes from another sector, then the corresponding liability resides in that originating sector so it can be booked as equity on the receiving sector's books.

9:28 AM  
Blogger Roman P. said...

Thank you both, you've driven me to read up on bank accounting and now I understand it a little bit more. I'm sorry if my answer to you is both late and sub-par; I've tried to write for most of the day and it's still sadly lacking both in its substance and presentation.

The assets the private banks have to exchange for the banknotes of Bank of Russia are monetary deposits on their correspondent accounts in the said Bank of Russia. Not sure if it is 100% equivalent to the Western system of reserve accounts, but apparently (after digging into the instructions for the bank accounting) it works about the same.

I now realize that I should have worded my thoughts better. Withdrawing cash from your bank account, of course, just shuffles already existing assets and liabilities (except for the private bank, which has to write them off). What I've meant by 'injections' was the injection of credit money by the private banks into the monetary circuit, not exchanging it into cash.

On the relationship of the NFA injections to the creation of the bank reserves I will have to read more papers to reach any conclusions.

We are probably speaking of different models. The one I've had in mind was the one from the paper "The Circuit Theory of Endogenous Money". There is no vertical channel in this model by design.

I don't think I could agree with your conceptual representation of the double entry bookkeeping. It is, in its nature, nothing more than a convention of how we should represent the stuff that we have. So when you say, 'this asset is balanced by this liability' - well, everything is balanced by definition. It's because double entry bookkeeping is an agreement to account both for what we have and how we got it. The liability side of a balance sheet represents the sources of the assets, and nothing more. Of course you are obliged to repay the money that came from the loan, but between that you could circulate received money in the economy and make them useful. The asset you received, money, is as good as if you received the transfer straight from the federal government, it's only the source (liability) that is arguably inferior one.

Also, it is simply wrong to say that one sector could not increase its equity without corresponding liabilities in another sector (you probably mean a government by it). Watch my clever trick: I decide to start a small steelworking factory. I need equity to start a firm (about a $300 in Russian law), but I do not need to supply it in cash. So I go to the field and with nothing but my hands and clay start to build a blast furnace. After I’m done, I call an expert and he says that because I’m so good, I’ve made a real and useful blast furnace worth said 300 bucks. Voila, I could start a firm with what I own. I fill out the forms, start to do the accounting… On the asset side of my balance sheet, I write down ‘Fixed asset – blast furnace ($300)’, and on the liability side I write ‘Owner’s contribution ($300)’. I have just created equity – and it is not matched by anything in the other sectors.

2:20 PM  
Blogger NeilW said...

If you inject credit money, you inject a credit asset as well - which net to zero.

And to inject credit money somebody has to want to borrow *first*. You can't force it upon the system. There is a finite demand for it at the current price.

"I have just created equity – and it is not matched by anything in the other sectors."

Correct in accounting terms. (although you omit the value of the field and where you got it from).

But that is real equity. You still have no way of settling your debts.

We settle debts in a monetary economy not by eliminating them but by swapping them for another acceptable debt.

1:23 AM  
Blogger Roman P. said...


I am a little afraid to contradict the man who corrected Keen's accounting in his models and presumably knows accounting far better than me, but I can't but disagree with the way you think of assets and liabilities. When you use a phrase "this and that net out to zero", there arises a confusion. Everything must net out to zero, really, just because for any asset there is a source. In that sense, if you found money on the street, or if you were gifted with them by someone, they will still 'net out to zero'. But they will not net out to zero in the same sense if you made an accounting posting and then immediately made an equivalent red storno posting.

I will argue that at a given time there is a disequilibrium: supply of credit is less than demand, especially now. But, so? The supply of credit money is endogenous, but when someone decides to get a credit and gets it, how is that any less of injection?

To settle my debt, I could produce some security backed by my asset (blast furnace). A promisory note, perhaps? You sell me pig iron, I give you the note. Then, I produce iron ingots and sell them back to you for some other security.

5:10 AM  
Blogger NeilW said...

I don't need iron ingots. So why am I going to advance you any credit on pig iron?

Anybody can create money. The trick is getting it accepted.

In a monetary economy we settle debts by swapping liabilities. And as the circuit study has shown that involves a third party acting as a bank

We don't have a barter economy. Trying to distill the system to one to win an argument gets us nowhere.

7:13 AM  
Blogger winterspeak said...

Roman: No, we have the same Keen model in mind. Yes, Keen has no vertical dimension by definition, and this is not accurate by definition.

And you are correct, double entry bookkeeping is just a convention on how to represent this, but it is true that any explanation needs to be consistent with the rules of double entry bookkeeping. There is a difference between money you have to pay back and money you don't, and the non-Govt sector has no way to increase the money it holds which does NOT have to pay back. If they want more of that, then you get paradox of thrift conditions and an output gap etc.

If you make something with your hands, you have created a real asset that you can then assign some nominal financial value to on the equity section of the liability side of the balance sheet. You are correct. However, I'm talking about financial assets, not real assets with a nominal value associated with them. You cannot settle a tax bill with a blast furnace, and you cannot convert your blast furnace to currency unless someone with currency is willing to trade, and this cannot happen at a sector level.

9:41 AM  

Post a Comment

Subscribe to Post Comments [Atom]

<< Home