Thursday, January 05, 2012

Stocks and Flows

A pithy summary of our macro debate by Brad DeLong:

The government purchases $100 billion of goods, issues $100 billion of bonds, and raises taxes by $3 billion a year in order to amortize the bonds. Government purchases go up by $100 billion this year. Private consumption goes down by $3 billion this year. Net fiscal impetus is not $0 but rather $97 billion. Cochrane doesn't understand the Ricardian Equivalence argument he is trying to make.

Now, ignore the Ricardian Equivalence, but just note that Keynesians think $97B was created via this spending and taxes this year, which is why they love government spending regardless of what it is spent upon. But if they issued $100B in bonds to create this 'new demand', where would that $100B have have gone otherwise? I don't think this makes sense in general equilibrium; money is never 'idle' unless it is under your mattress. I guess if you could caricature the two views: one thinks supply creates its own demand, the other that demand creates its own supply.


  1. Anonymous9:00 AM

    I thought the whole point was that people would see a big $100bln principal repayment looming in the future and therefore restrain spending to save for the future tax burden. Is this incorrect?

  2. Anonymous9:02 AM

    "where would that $100B have have gone otherwise?"

    Just wondering. What if the answer to this question were "chasing a bubble in government bonds". By bubble i mean a very narrow one (i.e. not one that would extend to other issuers)

  3. The main point is, as anonymous says, that people react to the $100 billion future tax increase (or spending cut, or reduction in value of nominal assets from inflation, or some other loss). You might argue that people will under-react in some cases, but it's highly implausible that they don't react at all, or that they systematically under-react (and there's no data supporting either implausible contention). I think systematic over-reaction is plausible (although also has no empirical evidence) since once currency debasement begins is almost always seems to accelerate.

    The question of where the $100 billion would have gone otherwise is a different one. It could come from private consumption or private investment, in either case likely making the net effect of the stimulus spending negative, even before factoring in the future costs. Government takes money today from privately-selected uses to government-directed ones (loss of utility there) and also must take money tomorrow from privately-selected uses to repay the debt.

    The Keynesian hope is that today's $100 billion comes from hoarding or asset bubbles, in which case the net effect could be neutral or even positive. The further hope would be that the future debt repayment will go to sound private investment or elimination of future wasteful spending (sort of "stuff then starve the beast").

    However "hoarding" and "bubble" are in the eye of the beholder. So even if you make the assumption people under-react to the future implications of stimulus spending, you also have to assume that the government's judgment using other people's money, with officials being paid whether they are right or wrong, is better than people making choices with their own money, bearing the losses if they are wrong.

    If people are correctly preparing for future bad times instead of "hoarding" or correctly anticipating a rise in nominal asset prices instead of feeding a "bubble", then the stimulus will be doubly harmful.

  4. Aaron: I guess I'm having trouble seeing idle balances as truly idle. Reminds me of the debates about corporations or banks sitting on too much cash, ergo we should take (tax) them on these idle balances. Big part of the 1937 recession, quickly rescinded.

  5. Anonymous10:43 AM

    Thanks for clarifying the first point, Aaron. I think what these guys miss with their $97 bin is that there are a whole stream of payments that must be taken into account (for example, taxpayers looking at their future liabilities). They are looking at one year only.

  6. I agree with you, I'm only trying to describe what I think Keynesians have in mind.

    We could imagine lots of people stuffing cash under their mattresses. I think they would do that for a reason, perhaps they expected deflation and bank failures. If the government took their money, they would cut back even more on consumption and investment to replenish their hoard. Also, I think other people are aware of the cash hoards, if the government took it away, those people would plan to produce fewer goods in the future. But a Keynesian might argue that people hoard for irrational reasons and if the government took the hoard away, hoarders would continue to spend and invest, and others would continue to produce, as if that money were still there.

    Similarly, people pay down debt when they fear bad times, which is reverse unhoarding and has the same economic effect as hoarding. If the government taxed them a Keynesian might hope that the people take out new debt to pay the taxes. I think it more likely that people would cut spending and investment even more.

    When people fear stagflation, they hoard real assets instead of cash. This soaks up value in the form of higher asset prices. If the government taxes people, Keynesians might hope people sell their asset hoards, reducing asset prices. I think people are likely instead to cut back on consumption and investment rather than part with their hard assets. Keeping these assets out of productive activity reduces production as well.

  7. But if they issued $100B in bonds to create this 'new demand', where would that $100B have have gone otherwise?

    I think DeLong answers this in his post:

    The Federal Reserve has raised the monetary base from $1 trillion to $3 trillion. But banks and firms haven't tried to spend their extra $2 trillion in liquid high-powered money balances but instead, "pathologically" says Cochrane, sit on them. Expansionary fiscal policy is a way of getting banks and firms to spend their cash balances buying Treasuries, and the government then gets to use those spent cash balances to hire people to do useful things.

    If I'm a bank and I sell a security to the Fed (QE), my reserves balance (held at the Fed) goes up. If I don't lend out those reserves, they never enter the economy and don't lead to economic activity. It's the same as if the Fed purchased a security from a private citizen and he/she put the proceeds under a mattress.

  8. Well, that's Monetary policy, not fiscal. With Monetary policy, money is created, creating nominal demand ex nihilo. With fiscal policy, it is not. If government borrows $100B to spend $100B, it seems a wash to me right away, unless that $100B was somehow not being used, but that's a rather antiquated view of money (ie, it is being horded outside of the investment-savings nexus).

  9. I bring up the QE example to illustrate what many perceive to be reality: banks are in fact hoarding reserves. The newly created $2 trillion has not found its way into circulation. That's the point.

  10. OK, banks think they need a large cash cushion, which given all the carping about excessive leverage, seems to be exactly what many commentators want. That's the current equilibrium. Taking $X B away from them and giving it to government just makes them hoard another $X billion. The new government borrowing would not add to demand, unless you are merely printing money to buy the bonds (or having the fed buy them, which is the same thing).

  11. Yes, but interest rates near zero blur the line between cash and Treasuries. Banks/other hoarders are looking for safety; the idea is you can give them Treasuries in exchange for cash and the government can then spend the cash.

  12. "If government borrows $100B to spend $100B, it seems a wash to me right away, unless that $100B was somehow not being used, but that's a rather antiquated view of money (ie, it is being horded outside of the investment-savings nexus)."

    What if that $100B was going to be spent on foreign sovereign debt?

  13. sean: Well, here it gets tricky. No bank actually owns 'cash' other than reserves for transactional balances within their atms and branches. What we call 'cash' is really 'Cash and equivalents', which means their assets are already working for them in some way, so it's not like they can buy Tbills with greenbacks.

    Dave: If that $100B was going for (say) Japanese debt, then the Japanese have $100B, which either buys consumption or investment goods in the USA.

  14. I don't agree with that, but let's not get sidetracked by that disagreement.

    Let's try a different angle.

    Doesn't the mere existence of 'excess' reserves imply that banks are not making all their assets work (ie. are hoarding cash)?

    From Greg Mankiw in the NYT in 2009:

    In the late 19th century, the German economist Silvio Gesell argued for a tax on holding money. He was concerned that during times of financial stress, people hoard money rather than lend it. John Maynard Keynes approvingly cited the idea of a carrying tax on money. With banks now holding substantial excess reserves, Gesell’s concern about cash hoarding suddenly seems very modern.

    Modern or 'antiquated'???

  15. I agree Mankiw said that, but I think he's wrong (he is a new-Keynesian, after all). I don't think any extra cash (really: Tbills) buffer is something to be appropriated and used better by someone else.

    I agree that bank reticence is an amplifier, but it's not in terms of leakage, but rather, it's a symptom that banks are risk averse do to their negative vega (see here).

  16. Anonymous9:47 PM

    He thinks that USD and treasury bills are perfect substitutes when the interest rate is zero. There's no reason for them to be treated any differently on the bank's balance sheet since the market in t-bills is so liquid. Meaning, there is no crowding out.

    Not that it's true.

  17. Anonymous11:56 PM

    The banks are risk averse and hoarding in a way that income constrained (and formerly unemployed) individuals are not. So the government can increase the velocity of money simply by virtue of hiring unemployed people to do "stuff."

    Delong's blog has a new post explaining this concept.

  18. Anonymous2:23 AM

    "I don't think this makes sense in general equilibrium"

    Maybe the economy isn't in general equilibrium.

  19. Anonymous11:31 AM

    But Anon, even if the economy isn't in formal equilibrium, it certainly looks very stable. Economic "instability" is very relative - a couple of percentage points, at most. If the Delong argument was true, then why doesn't the economy see truly large oscillations - NGDP growing by an order of magnitude in a single year, for instance. Because you will note that nothing in Delong's argument is unique to the government - it would be the case whoever borrowed the money.

  20. Anonymous11:35 AM

    ^ True, but "nobody" wants to borrow money, they just want to park it in safe assets like, for example, US government debt. Leaving the US government as the borrower and spender of last resort.

    Households, corporations, and the government can't all delever at the same time.

  21. Good discussion in the comments. Fundamental question between Eric and Sean: are the banks holding excess reserves because the banks want the liquidity (suggesting an equilibrium), or because they don't want to lend (suggesting idle cash)?

    I would point out a few things. One, the Fed is paying interest on reserves, which may encourage excess reserves. Two, even if the banks are sitting on idle cash, this does not necessitate fiscal stimulus. The banks can buy assets other than Treasuries from markets or institutions that aren't hoarding. AAA Corporates, for instance. Monetary policy would then accomplish the same increase in velocity as the bond issuance for fiscal policy.

  22. question: if the banks are holding excess reserves shouldn't that be reflected in higher equity values for banks?

  23. This discussion is getting off track due to a common misconception about excess reserves. The size of the monetary base (reserves plus currency) is determined entirely by the Fed. They cannot, in aggregate, be changed by the banking system. When one bank spends it's excess reserves on a security or loses a deposit that excess reserve gets transferred to another bank. The evidence for the Keynsian claim that there is insufficient aggregate demand is 1) there is large scale unemployment 2) govt and other high quality debt has gone up a lot in price despite increasing in supply, suggesting lots of demand for riskless assets. That implies that the private sector is leaving real resources unallocated and they view fiscal policy as able to use those real resources productively. Fiscal stimulus does not deprive the private sector of their desired cash or riskless safety reserves. In fact, it satiates them by creating riskless assets the private sector desires (govt debt issuance) and mobilizes the unused real resources. In the future, when the private sector no longer desires such a large stock of riskless assets, the govt will reverse the process and release the real resources to the private sector and pay down its debt. This process is meant to counteract the paradox of thrift that underlies an aggregate demand based recession.

  24. I forgot to add that, in the Keynsian framework, this fiscal stimulus is only necessary when interest rates are at the zero bound. When rates are substantially positive (i.e. when the economy is in "equilibrium")and the private sector desires to build a cushion of reserves the price of risk free assets will rise to the point where the private sector no longer wishes to aquire them and full employment will be maintained. Fiscal stimulus only becomes necessary when rates reach the zero bound and interest rates cannot fall far enough to stop a paradox of thrift.