It’s stupid to assume stupidity. Er, I mean it’s not wise to assume stupidity

Here’s Paul Krugman:

Menzie Chinn has some fun pointing out that if the doctrine Heritage was pushing to oppose fiscal stimulus were true — namely, that government borrowing always crowds out an equal amount of private spending — then the fiscal cliff could not be a problem.

.  .  .

But what Menzie doesn’t mention is that the very same doctrine was propounded by distinguished economists at the University of Chicago — John Cochrane and Gene Fama made exactly the same argument that Brian Riedl was making at Heritage, while Robert Lucas fell into a somewhat different but equally misleading fallacy.

So if you think the fiscal cliff matters, you also, whether you know it or not, believe that a whole school of macroeconomics responded to the greatest economic crisis since the Great Depression with ludicrous conceptual errors, of a kind nobody has had a right to make since 1936 at the latest.

Or perhaps one should take a deep breath, and spend a minute or two recalling that conservatives think the supply-side effects of taxes are very important.

PS.  In the post on Lucas that Krugman links to, he quotes Lucas as saying:

If the government builds a bridge, and then the Fed prints up some money to pay the bridge builders, that’s just a monetary policy. We don’t need the bridge to do that. We can print up the same amount of money and buy anything with it. So, the only part of the stimulus package that’s stimulating is the monetary part.

But, if we do build the bridge by taking tax money away from somebody else, and using that to pay the bridge builder — the guys who work on the bridge — then it’s just a wash. It has no first-starter effect.

Since Lucas was my adviser I won’t call this ludicrous, but it is wrong.  He’s saying that it has no effect holding M constant, when he should say holding M*V constant. Or it would have no effect if the central bank were targeting inflation, or the Taylor Rule, or whatever.  What he’s really saying is it makes no sense to do fiscal stimulus if the monetary authority is targeting some sort of nominal aggregate like M*V.  And that if fiscal stimulus works, it’s just a backdoor way of doing monetary stimulus (i.e. more M*V.)  That’s all true.

Here’s Krugman’s response:

I’ve tried to explain why Lucas and those with similar views are all wrong several times, for example here. But it just occurred to me that there may be an even more intuitive way to see just how wrong this is: think about what happens when a family buys a house with a 30-year mortgage.

Suppose that the family takes out a $100,000 home loan (I know, it’s hard to find houses that cheap, but I just want a round number). If the house is newly built, that’s $100,000 of spending that takes place in the economy. But the family has also taken on debt, and will presumably spend less because it knows that it has to pay off that debt.

But the debt won’t be paid off all at once — and there’s no reason to expect the family to cut its spending right now by $100,000. Its annual mortgage payment will be something like $6,000, so maybe you would expect a fall in spending by $6000; that offsets only a small fraction of the debt-financed purchase.

Now notice that this family is very much like the representative household in a Ricardian equivalence economy, reacting to a deficit financed infrastructure project like Lucas’s bridge; in this case the household really does know that today’s spending will reduce its future disposable income. And even so, its reaction involves very little offset to the initial spending.

How could anyone who thought about this for even a minute — let alone someone with an economics training — get this wrong? And yet as far as I can tell almost everyone on the freshwater side of this divide did get it wrong, and has yet to acknowledge the error.

I won’t call this ludicrous, but it’s a very strange argument.  Krugman should have said that Lucas forgot that fiscal stimulus can impact V.  But instead he makes a completely different argument that doesn’t seem to depend on V rising; one that opens the door to deflationary stimulus, i.e. real GDP going up without any rise in either M or V.  That’s possible, but not much of a refutation of Lucas.

I also don’t understand the representative household argument.  If the stimulus is to work, it must raise the quantity of labor supplied by the representative household.  If Krugman had made the standard Keynesian argument that more fiscal stimulus boosts NGDP, then he’d just need to assume sticky wages to get more hours worked.  But here he seems to be making an optimal consumption smoothing argument that requires people to work more hours, but doesn’t explain why they would want to work more hours.  Is it because the bridge makes them poorer?  It’s obvious to me that the representative household that buys a $100,000 house will not reduce consumption by an equal amount.  But it’s not obvious that they would avoid that drop in consumption by choosing to work more.

So what’s the Keynesian model all about?  Is it about nominal shocks combined with sticky wages and prices, or is it about inter-temporal consumption allocation decisions when the government decides to build bridges?

And if it’s “obvious” the bridge will make real GDP rise, is it equally obvious it will work if the central bank targets NGDP, or does it only work if the central bank targets M?  If it’s the latter, then Krugman should have just claimed that bridges boost velocity, and left it at that.  If the former, then he’s claiming bridges can provide deflationary growth.  Maybe, but that’s not the Keynesian argument, is it?

I wish Keynesians and freshwater economists could agree on a common language:

1.  There are nominal shocks when NGDP changes, and they may have real effects if wages and prices are sticky.

2.  There are all sorts of real shocks that can also impact RGDP.

Instead freshwater economists spend so little time thinking about nominal shocks that they forget it’s M*V that matters, not M.  And Keynesians spend so little time thinking about real shocks that they don’t see how fiscal policy could affect anything beyond AD.

Update:  Marcus Nunes got there first.


Tags:

 
 
 

49 Responses to “It’s stupid to assume stupidity. Er, I mean it’s not wise to assume stupidity”

  1. Gravatar of Rien Huizer Rien Huizer
    9. December 2012 at 10:16

    Scott,

    Why do such distinguished economists make such obvious mistakes? Are they irrational?

  2. Gravatar of marcus nunes marcus nunes
    9. December 2012 at 10:29

    Scott,
    Short version:
    http://thefaintofheart.wordpress.com/2012/12/08/keynes-krugman-and-the-fiscal-cliff/

  3. Gravatar of Saturos Saturos
    9. December 2012 at 10:34

    Krugman appears to be confusing Ricardian equivalence with the more basic “where does the money come from” problem that Lucas was raising there. Indeed Lucas goes on to say, “And we know that borrowing the money instead won’t help” – that’s the Ricardian bit. Few economists clearly understand that fiscal stimulus only works: a) at the ZLB, when there is a glut of loanable funds, or b) at normal times, by raising rates (and thus reducing the demand for money) and also by transfering income from the public to the government which has less demand for holding money.

  4. Gravatar of marcus nunes marcus nunes
    9. December 2012 at 10:34

    “Stupidity” seems to be “getting around”:
    http://thefaintofheart.wordpress.com/2012/12/08/6323/

  5. Gravatar of Ram Ram
    9. December 2012 at 11:24

    Krugman’s post about the mortgage is a bit difficult to understand, but is it possible that he’s saying something like the following?

    “Lucas thinks that if the government spends some money and boosts V, holding M constant, then the soon to be taxed households reduce V by the same amount. My example suggests that the expectation of future taxes lowers V by less than the government spending raises V.”

    This is not my view, in any case, but it seems like the most charitable reading of his supposed refutation in equation of exchange language. Needless to say, even if he were right, that wouldn’t make Lucas’s argument ludicrous.

  6. Gravatar of Kevin Donoghue Kevin Donoghue
    9. December 2012 at 11:54

    Scott: If the stimulus is to work, it must raise the quantity of labor supplied by the representative household.

    Is that what you actually meant to write? If you had written “the quantity of labor sold” you’d be right, obviously. But where there is involuntary unemployment, quantity sold is less than quantity supplied. The household is rationed in the labour market.

    No big deal if you were just slipping into layman’s terminology. I only mention it because, honestly and all snark aside, I often get the sense that your evident contempt for Keynesian thinking has prevented you from actually getting to grips with Keynesian models (which is not something that I see in Nick Rowe’s blogging, for example).

  7. Gravatar of AA AA
    9. December 2012 at 12:40

    If I understand correctly, fiscal policy raises V, so if we assume M is unchanged, then with sticky prices NGDP and RGDP will increase?

  8. Gravatar of Major_Freedom Major_Freedom
    9. December 2012 at 13:41

    Krugman writes:

    But the debt won’t be paid off all at once — and there’s no reason to expect the family to cut its spending right now by $100,000. Its annual mortgage payment will be something like $6,000, so maybe you would expect a fall in spending by $6000; that offsets only a small fraction of the debt-financed purchase.

    This argument is severely flawed. It seems the family’s reduction of $6,000 in “spending” somehow disappears from the economic system. But the lender is getting $6,000 in additional income each year, so wouldn’t that enable the lender to put to use $6,000 more each year going forward? Why should spending as such fall?

  9. Gravatar of CA CA
    9. December 2012 at 14:20

    Mary Matalin and George Will got a couple good zingers in on Krugman this morning.

    http://abcnews.go.com/ThisWeek/video/roundtable-fiscal-cliff-stalemate-17918526?tab=9482930&section=1206874&playlist=17919121

  10. Gravatar of Greg Ransom Greg Ransom
    9. December 2012 at 14:49

    “It’s stupid to assume stupidity.”

    Isn’t this the whole basis of you endless posts on Sheldon Richman?

  11. Gravatar of ssumner ssumner
    9. December 2012 at 15:07

    Rien, Everyone makes mistakes, especially when speaking off the cuff, as Lucas was doing. I’d hate for people to write down what I said on economics in casual conversation. Cochrane’s problem is that he speaks another language from Keynesianism.

    Ram, It’s not my area, but it reminds me of “real” arguments for fiscal stimulus that I have seen. The basic argument is that fiscal stimulus makes people feel poorer, so they work harder. But I can’t imagine that’s what Krugman intended, so I am probably misinterpreting his argument. Interestingly, Wren-Lewis made a similar argument that was erroneous. But I don’t see any logical flaw in Krugman’s argument.

    AA, Yes, it will probably increase V, although the amount by which V rises has little to do with intertemporal consumption decisions. Whether it raises M*V depends on what the Fed’s targeting. BTW, in the PS I wasn’t claiming Krugman was wrong, just that he was making a very peculiar argument in trying to show Lucas was wrong. Unless I am mistaken, he wasn’t making a standard Keynesian argument, which is that fiscal policy only works by shifting the AD curve to the right. His mechanism would seem to work even if NGDP was constant.

    Kevin, In all the economics books I’ve ever read “supply” refers to the schedule of desired sales at various price points, and “quantity supplied” is the actual quantity sold. But then maybe you’ve read other economic books.

    Supply is not a number, it’s a schedule. Quantity supplied is a specific number.

  12. Gravatar of ssumner ssumner
    9. December 2012 at 15:09

    CA, You have to give me the specific time, or I won’t watch.

    Marcus, We think alike. I added a link.

  13. Gravatar of Steve Steve
    9. December 2012 at 16:24

    In a nutshell, Scott sums up thousands of hours of debate:

    “Instead freshwater economists spend so little time thinking about nominal shocks that they forget it’s M*V that matters, not M. And Keynesians spend so little time thinking about real shocks that they don’t see how fiscal policy could affect anything beyond AD.”

  14. Gravatar of Saturos Saturos
    9. December 2012 at 18:46

    I was kind of hoping you’d get to this in your own time, but could you please comment on this Tyler post? Once again, I assume your standing counterargument to Tyler’s views is 1933?

    http://marginalrevolution.com/marginalrevolution/2012/12/what-the-beckworth-diagram-really-means.html

  15. Gravatar of Tommy Dorsett Tommy Dorsett
    9. December 2012 at 19:27

    I don’t see any evidence that a larger cyclically adjusted fiscal deficit raises the velocity of money:

    http://research.stlouisfed.org/fredgraph.png?g=dER

    In fact, the relationship usually runs the other way.

    Maybe Lucas was right after all.

  16. Gravatar of Suvy Suvy
    9. December 2012 at 19:32

    I think that Krugman misses the central point about the effectiveness of fiscal policy. Monetary policy can only be used to buy financial assets; it cannot be used to hire people and increase production. The role of fiscal policy can be to put some portion of the 12 million people out of work to produce something valuable. However, you have to be careful not to increase the public debt too much as a result of these policies as too much public debt is also counterproductive to growth. In most recessions, fiscal policy is absolutely useless. However, when you’re stuck in a depression with an extended period of high unemployment and slow real growth, fiscal policy can make sense provided that it is used to make something productive that can provide a return like infrastructure or public works.

    The tricky part about fiscal policy is that you’re leaving it to the discretion of politicians. That’s the part that I really dislike about it. However, I think it can be used well if it is done in a decentralized manner. Rather than having the federal government allocate the resources, it might be better to give the money directly to state and local governments and let them allocate the resources.

    Another thing I hate is when politicians run deficits in the boom times. It makes to sense and fiscal policy should not be used during the boom times. It is counterproductive as it just takes away resources that the private sector could use more effectively. We need to run surpluses during the boom times to pay for the fiscal policy we use during the busts.

  17. Gravatar of jknarr jknarr
    9. December 2012 at 21:20

    I find Krugman a very slippery and dishonest writer, largely because he introduces his (deeply flawed, typically close to the edge socialist) premises with brief, moderate, and reasonable language, and later emotionally distracts by calling people stupid. It’s like missing a magicians sleigh of hand, because the beautiful assistant is cursing you and calling you an idiot. Calling people stupid is key to his rhetorical model — raw brains for his zombie army, and lizard-brain-level thinking for his critics. Given his track record of this in the NYT, I am frankly amazed that he is still listened to. Entertainment, perhaps, but really not to be taken seriously, surely — Maureen Dowd in drag.

    Shouldn’t we be talking about how the saving of $100,000 in the first place required a saver and deferred consumption in the first place? In short, it took an asset surplus over liabilities.

    I saved that volume of capital, I could buy a Ferrari or make this loan. I choose to loan, and therefore no Ferrari, no additional capacity to lend to another guy — the funds are tied up. He assures us that the household does not cut by $100,000?! This guy is shameless.

    Leverage changes this all, but is confused with capital/equity. Perhaps the other side of the house trade now has $100,000 on deposit in the bank to buy that Ferrari because he had tons of equity; or perhaps he needs to pay off a $100,000 loan himself — and hands it to the bank, who shifts required- into excess- reserves, and the “money” is gone — poof. In other words, the chain of balance sheet qualities — leverage — determines whether the funds are available for others to borrow, or if it simply extinguishes a debt chain.

    Capital markets exist because there is capital — a surplus of assets over liabilities. Thought experiment — what is potential V if there is only very little capital (high leverage)? What is potential V if all balance sheets are deleveraged (high savings/capital)? Might V simply be the marginal propensity to leverage — which hits a limit? Just as Keynes pointed out that there is a limit to savings — because it is derived from income — can there be a limit to dissavings?

    If high leverage collapses V, an injection of base money would boost nominal cash flows, and crush real debt burdens — deleveraging, and increasing potential V. V might then only make sense if one differentiates clearly between base- money and all its debt derivations (higher Ms). We appear to constantly fail to define what “the money” is.

  18. Gravatar of Steve Steve
    9. December 2012 at 23:09

    Krugman has a new piece up: “The Oppression of the Monopolistic Robots” or something. I guess “Neo-Krugman” is trying to become the Keanu Reeves of the economics profession.

    I don’t have time to dig through the raw data in his chart, but I’m suspicious it might actually support NGDP targeting. That’s because labor share of output declines a few percent, while “depreciation is a rising share of the total.” I’m really not sure how this data series is constructed, but depreciation schedules tend to be fixed over time, so if nominal labor income fell >10% below trend vs. stable depreciation, that would be enough to knock his series several percentage points below trend.

    Anyone care to investigate?

  19. Gravatar of Kevin Donoghue Kevin Donoghue
    10. December 2012 at 02:16

    Scott writes: In all the economics books I’ve ever read “supply” refers to the schedule of desired sales at various price points, and “quantity supplied” is the actual quantity sold. But then maybe you’ve read other economic books.

    Indeed I have. Malinvaud, The Theory of Unemployment Reconsiderd is the obvious reference. If a seller is rationed “his sale is smaller than his supply” (Chapter 1, section 4).

    It’s true to say, as you do, that quantity supplied is a specific number; but so is quantity sold, which may be a smaller number.

    I won’t be shocked if you tell me, as you did in another thread, that you have no idea what I am talking about. I can only repeat what I said there, that I’m expressing myself as simply as I can.

  20. Gravatar of Saturos Saturos
    10. December 2012 at 03:46

    A question from Justin Wolfers: https://twitter.com/justinwolfers/status/277050083252510721

    Kevin, I agree.

  21. Gravatar of Morgan Warstler Morgan Warstler
    10. December 2012 at 03:56

    Ok everybody, grab hold of your jocks.

    I “think” DeKrugman and I arre about to agree.

    BUT, it is my policy proposal that will expose his weakness for you all to jump on.

    In past week DeKrugman 3.0 has been unveiled.

    http://www.nytimes.com/2012/12/10/opinion/krugman-robots-and-robber-barons.html

    Herein, he goes to 11, and is clear as a bell about robber barons.

    cont…

  22. Gravatar of Saturos Saturos
    10. December 2012 at 03:58

    And Scott, what do you think of this argument?

    https://twitter.com/IanShepherdson/status/276746370508005376

  23. Gravatar of Morgan Warstler Morgan Warstler
    10. December 2012 at 04:08

    Since you’ve known me, I have been an advocate of what I call encoded smallness.

    I favor simple and easy, low hanging fruit, policies which favor SMBs. I view my position as a non-catholic modern day distributism.

    Problems with capitalism ONLY arise when there are not enough capitalists.

    The Chuck Norris argument is:

    Big Biz= Big Gov 100% of the time.

    None of you, no one, not ANYONE has ever had the nuts to argue this with me.

    —–

    So even though in the short term, laws which favor smallness, appear to be “inefficient” they are actually like the US Constitution – in the long term the BEST WAY OF SURVIVING.

    This is why Glass-Steagall, was a good law, even though blahblahblah…

    It was simple. It was encoded smallness.

    The thing is this, encoded smallness has NOTHING to do with labor.

    The DeKrugman argument on labor is bullshit.

    BUT, since he is so hotted up about robber barons, we can use “encoded smallness” as a response to his complaint.

    He will NOT favor encoded smallness.

    Because Big Biz = Big Gov

    and he FORGOT that.

    Eventually Krugman will cheer for Big Biz > SMB, since it promotes Big Gov.

    But right now, he’s exposed.

  24. Gravatar of Jason Jason
    10. December 2012 at 04:09

    I think, as a rule, if you are trying to say something sophisticated, then don’t act like its something simple.

    If you are saying something complicated about the interaction of monetary and fiscal policy, don’t act likes its simply the we live in a world where G is always crowded out by decreased I and C.

  25. Gravatar of Y.Alekseyev Y.Alekseyev
    10. December 2012 at 04:16

    Scott – a good and lucid post. Thanks!

    @Tommy Dorset: So? That graph tells you nothing because V falls when households deleverage, and the increase in government borrowing may or may not fully offset it (usually does not). Very rarely, if ever, was expansion in government
    spending so large as to MORE than offset private sector deleveraging. WWII may be the exception.

    @jknarr: the question of “where did the 100,000″ come from is only relevant if you think only M matters. But see Scott’s post above. MV is the variable you want looking at.

  26. Gravatar of Benny Lava Benny Lava
    10. December 2012 at 06:10

    I would just like to point out:
    -at current rates buying a house for 100,000 would give you a monthly bill far less than 6k. More like 1k depending on insurance, your FICO, etc.
    -the hypothetical family would probably be moving from a rental. So in the long run they will probably save money via fixed rates and a declining principle.

    I don’t think it a well crafted example but I think I understand the meaning. The family will in the long run spend less but the short run consumption rises. Which probably isn’t true in reality but yeah.

  27. Gravatar of Ritwik Ritwik
    10. December 2012 at 06:25

    Saturos

    Your argument on fiscal policy (comment 3) is bang on, and I’d even say superior to the standard MIT-Keynesian narrative. Not at all bad for a card carrying market monetarist!

    I’d just add the next step from the argument about the relative demand to hold money between ‘the public’ and the government – the relative demand to hold money between the wealth-preservers and the spenders (including investors). Arguments about the embedded fiscal-ness and non-neutral distributive nature of most standard conceptions of monetary policy follow.

    I’m more and more convinced that at least in terms of monetary actions that affect the CB’s balance sheet, the benchmark case of a pure, fiscally-neutral injection of money should be a helicopter drop. All other balance sheet policy is fiscally non-neutral, both in the aggregate as well as in distributive consequences.

  28. Gravatar of ssumner ssumner
    10. December 2012 at 07:01

    Saturos, I mostly agree with Tyler, although of course I’m a bit more optimistic on what NGDPLT could have done in 2008-09.

    Tommy, That’s tough to say, as big deficits tend to occur during recessions, when other factors are depressing V.

    Kevin, You said;

    “Indeed I have. Malinvaud, The Theory of Unemployment Reconsiderd is the obvious reference. If a seller is rationed “his sale is smaller than his supply” (Chapter 1, section 4).”

    I agree, that’s why I used the term ‘quantity supplied’, and not supply. In any case it’s a moot point even if I am wrong, because in context the meaning was obvious.

  29. Gravatar of ssumner ssumner
    10. December 2012 at 07:05

    Saturos, I don’t trust either RGDP or employment numbers, at least in the short run. The longer term trends are somewhat better.

    The 45% MTR post is just silly. The US has a higher percapita GDP than Australia (PPP adjusted.) Growth rates are not affected by MTRs, levels are. I see people make this levels/growth rates mistake all the time, especially when they tout the “Chinese model” which has produced a country much poorer than Mexico.

  30. Gravatar of ssumner ssumner
    10. December 2012 at 07:23

    Kevin, I second thought I think you are right. Brain freeze.

  31. Gravatar of Saturos Saturos
    10. December 2012 at 07:49

    “Growth rates are not affected by MTRs, levels are.”

    I think there is something of a growth effect in the long run, due to human capital and such. But that was my thinking too:

    https://twitter.com/_Srijit/status/278162489353572353

  32. Gravatar of Saturos Saturos
    10. December 2012 at 07:55

    Scott and Kevin, yep, just think price floors.

  33. Gravatar of Kevin Donoghue Kevin Donoghue
    10. December 2012 at 08:53

    Saturos/Scott, thanks. So to get back to Scott’s puzzlement with Krugman, with ‘sold’ in place of supplied:

    I also don’t understand the representative household argument. If the stimulus is to work, it must raise the quantity of labor [sold] by the representative household. If Krugman had made the standard Keynesian argument that more fiscal stimulus boosts NGDP, then he’d just need to assume sticky wages to get more hours worked. But here he seems to be making an optimal consumption smoothing argument that requires people to work more hours, but doesn’t explain why they would want to work more hours.

    Krugman is assuming that the household wants to sell more labour at the current price, but demand is deficient. Even though the deficit-financed building of Lucas’s bridge reduces the present value of the household’s lifetime income, consumption demand is scarcely affected at all.

    Krugman isn’t wandering off the reservation here by any means. AFAICR this argument cropped up in Robert Barro’s writings in the 1980s, where he concluded that the multiplier for bond-financed government purchases equals the balanced-budget multiplier (and equals 1 if there is involuntary unemployment). It’s one of the ironies of macro that, in his youth, Barro wrote some very clear expositions of Keynesian economics.

  34. Gravatar of Kevin Donoghue Kevin Donoghue
    10. December 2012 at 09:00

    Let me just amend one sentence which maybe isn’t so clear:

    Even though the deficit-financed building of Lucas’s bridge reduces the present value of the household’s lifetime (after-tax) income, consumption demand (in the current period) is scarcely affected at all.

  35. Gravatar of Doug M Doug M
    10. December 2012 at 09:07

    When a person borrows 100,000 to buy an new house, money is created and demand increases in the short-term. Possibly, there would be a small reduction in demand in the following year, as that consumer is not buying a house that year. That is the way that GDP accounting works, a one of spike in demand or investment is growth this year and negative growth the following year. Furthermore, there may be a long term drag on consumption as a fraction of the mortgagors income is now tied up in debt service.

  36. Gravatar of Kailer Kailer
    10. December 2012 at 09:20

    It seems to me there are two primary ways that the government can get me to give it money. It can threaten to put me in jail if I don’t give it money (taxes), or it can offer to pay me in the future for the money I give it today.

    I think a lot of Keynesians would agree that government spending that’s fully funded by current period taxes would be largely offset by reduced spending since the government took the money from people, so now they can’t spend it. I don’t see why deficit funded government spending is fundamentally different. Why does it matter whether the government got the money from me by using the carrot or the stick? Either way I can’t spend that money on anything else.

    I can think of a number of ways that deficit funded government spending could actually stimulate spending. Off the top of my head, it could be that the money the government borrows comes from people that would otherwise have just sat on the money, whereas taxes come from people that would otherwise have bought a bunch of stuff. But it’s a lot more nuanced than G goes up, so Y goes up, recession over.

  37. Gravatar of Floccina Floccina
    10. December 2012 at 09:52

    The fiscal cliff matters because the deltas are what hurt the economy. People are often Laid off faster than they can be hired by another employer.

  38. Gravatar of Morgan Warstler Morgan Warstler
    10. December 2012 at 10:31

    “It seems to me there are two primary ways that the government can get me to give it money. It can threaten to put me in jail if I don’t give it money (taxes), or it can offer to pay me in the future for the money I give it today.”

    you get this is really only one way, right?

    Once you remove the ponzi scheme, all that’s left is the robbery bit.

  39. Gravatar of J Mann J Mann
    10. December 2012 at 12:14

    My head hurts – it never pays to try to figure out what Krugman is saying. Here’s my reading of his parable.

    1) The household is a proxy for the US combined public and private economy. Obama is the head of household, private industry are the kids. In year 1 of the Krugman-Obama recovery plan, the government spends an extra $100,000* in fiscal stimulus over what it would have spent if it stuck to trend.

    *multiply as necessary until it fits.

    2) (Possibly) People realize that the government eventually going to have to recover this $100K in borrowed money, so they reduce their spending somewhat along the lines suggested by the lifetime earning hypothesis, which results in the $100K producing more economic activity (i.e., employing slack resources) today, at a cost of a slower growth rate in the future.

    3) (Or maybe) The government essentially monetizes this $100K by issuing $100K in bonds, which it sells to itself, using money it prints to pay itself. (I think). The central bank can’t or won’t counteract this, so the fiscal spending increases V and/or spurs inflation.

    Ok, I got lost somewhere before 2 and was groping around. It would be nice if Krugman just spelled out whatever it was he was trying to say.

  40. Gravatar of Bob Murphy Bob Murphy
    10. December 2012 at 15:22

    Scott,

    You know we are evenly matched in our disdain for Krugman, so my comment here is just to explain what I take his argument to be; I’m not saying he’s right in the grand scheme. But you seem genuinely puzzled, whereas to me it’s obvious what he’s saying here. So:

    1) Krugman thinks Lucas and others are getting the standard Ricardian Equivalence result wrong. Lucas et al. are jumping from “a deficit-financed tax cut won’t change anything real” to the similar but actually distinct assertion “a deficit-financed increase in government spending won’t change anything real.”

    2) To illustrate why this is wrong, Krugman realized he doesn’t need to get hip-deep into a New Keynesian model. He just needs to remind people that households don’t completely reduce their “total spending” when they buy a house via a mortgage. So, by the same token, even if everyone has Rational Expectations and plans on paying permanently higher taxes because of the government’s huge deficit this year, that won’t perfectly offset the stimulus. In fact, it won’t even come close. So Lucas et al. are totally misapplying Ricardian Equivalence-ish arguments when they claim that a $1 trillion deficit-financed stimulus will just be a wash.

    3) You ask why this would increase employment. Well, for *whatever* reason increased demand would boost employment. Krugman thinks he is showing that the government can boost demand, and that even Perfectly Rational private sector won’t perfectly offset it. Hence, higher demand and hence higher employment, if we originally started out with idle resources.

  41. Gravatar of Kevin Donoghue Kevin Donoghue
    10. December 2012 at 15:39

    Bob,

    FWIW, I reckon you get it, if by ‘deficit financed’ you mean ‘bond-financed’ as I assume you do. Incidentally, Simon Wren-Lewis was saying much the same thing as Krugman.

    Just what Lucas was really trying to say I don’t know. But if you grant that Lucas was indeed making the mistake he appeared to be making, then Krugman’s response is correct.

  42. Gravatar of ssumner ssumner
    10. December 2012 at 18:30

    Bob, You may be right, but that’s not really a Keynesian argument. It’s an argument that works equally well in new Classical models. I’ve seen Barro make this argument.

    But in in a new classical model you also have to explain why people work harder, and obviously you CANNOT use the standard assumption that the combination of more AD and sticky wages increases employment. You actually must make people want to work harder. That’s why I talked about quantity of labor supplied going up. One method is to assume that the public believes the bridge makes them poorer–they get no utility from it. In that case they may work harder. Perhaps there are other ways of doing this as well. But the way Krugman described it didn’t seem Keynesian at all.

    BTW, It’s possible that what Lucas was thinking is that there is no “multiplier” in the sense of a number bigger than 1.0. Thus in the Krugman example the multplier might be 0.3. Obviously that’s positive and as I said Lucas is technically wrong, but I’d guess he was thinking in terms of secondary effects. But on the policy question Lucas was obviously right–we should just do monetary stimulus, fiscal stimulus is a giant waste of money. And that’s also obviously the point he was trying to make, but made it in an incorrect way.

    Kevin, All of these arguments need to distinguish between nominal and real effects. I can imagine Krugman’s argument working even if you had perfect wage price flexibility and no involuntary unemployment. That’s why I thought it was an odd argument to use.

    There are other Keynesian arguments that depend on fiscal stimulus being able to boost NGDP, and also there being involuntary unemployment due to wage and price stickiness. But the way Krugman worded his argument he seemed to be making the sort of argument I associate with Barro. Which is fine, but I think it would have been easier to simply say:

    1. Lucas forgot that V will rise.

    or

    2. The bridge makes people feel poorer, so they WANT to work harder.

    Instead he constructs it in such a way that it seems like it’s a Keynesian argument, but it really isn’t.

    And here’s another way of describing my puzzlement. Suppose you lived in a country where everyone liked to hold 10% of their nominal income as base money. Now the first mechanism won’t work, but the second one will. Is the second one “Keynesian economics?” I suppose that’s debatable.

  43. Gravatar of Bob Murphy Bob Murphy
    10. December 2012 at 20:52

    Scott,

    I still don’t quite see your problem. Again, I think Krugman is wrong, but only because I don’t buy his model. You seem to think he is misapplying his model somehow? Here’s the progression:

    (1) Krugman thinks wages are sticky and we’re in the liquidity trap. Sure, unconventional monetary policy could work, but Bernanke can’t do the right thing because of Ron Paul’s goon squad. So we’re left with fiscal policy. Krugman recommends the government borrows an extra $1 trillion to build a bridge to Palin’s house.

    (2) Krugman *interprets* Lucas, Fama, et al. as saying, “No, that won’t boost demand at all. Taxpayers are rational, and know they will pay higher taxes to finance the bigger debt because of this bridge. So private spending will fall by exactly $1 trillion. Thus no jobs are created.”

    (3) Krugman says, “That is the dumbest thing I’ve ever heard. I don’t even need a fancy Keynesian model, just think about someone buying a house with a mortgage.”

    Does the above make sense? That’s as sophisticated at Krugman is being. You keep asking, “Why do people work harder?” but the answer is, because there is insufficient demand. I could ask why *you* think people will work harder once Bernanke takes your advice. Do people think targeting NGDP makes them poorer? (Well *I* think that, but apparently you don’t, and you don’t need it to make your model work…)

  44. Gravatar of Kevin Donoghue Kevin Donoghue
    11. December 2012 at 01:54

    Suppose you lived in a country where everyone liked to hold 10% of their nominal income as base money.

    I see no problem for Keynesian economics in such a world. In fact when I try to work out a proper formal model with an explicit utility function, I typically assume:

    U = h.LogC + k.Log(M/P) with t subscripts if it’s T>1 period model.

    In such a model velocity is always constant (or at least I haven’t found a case where it isn’t). But if there is involuntary unemployment (or idle resources as Bob Murphy prefers to put it) then the multiplier is positive.

    Bob is right. You’re making this more complicated than it is. Also, you’re reinforcing my suspicion that I mentioned in my first comment: your contempt for Keynesian thinking has prevented you from actually getting to grips with Keynesian models. That’s the drawback of Chicago training I guess. Bob shares your contempt, but he does seem to have done the homework.

  45. Gravatar of ssumner ssumner
    12. December 2012 at 05:10

    Bob, I think people will work harder because wages and prices are sticky. But the mechanism Krugman discusses is a “real” mechanism, it doesn’t require any wage and price stickiness to get people to work harder.

    Here’s another weay of putting things. If Lucas had talked about MV (as he should have) instead of M, Krugman could still have used the same argument.

    Kevin, I certainly think I understand what Bob is saying, I’m not so sure you and Bob understand what I am saying. You do not need involuntary unemployment for Krugman’s argument to get more output–lots of people have published articles showing that.

    When the government builds the bridge people will WANT to work more, because if they don’t their consumption will have to fall, and people like to smooth consumption. No involuntary unemployment is needed.

  46. Gravatar of Kevin Donoghue Kevin Donoghue
    12. December 2012 at 05:53

    Scott, maybe it would help if I simply list some statements of yours which look wrong to me:

    Krugman should have said that Lucas forgot that fiscal stimulus can impact V.

    Fiscal stimulus will work, on Krugman’s assumptions, even if it leaves V unchanged. Consider the U(C, M/P) function in my previous comment.

    If the stimulus is to work, it must raise the quantity of labor supplied by the representative household.

    No, it must raise the demand for goods and hence the demand for labour. The household is currently selling less labour than it desires to.

    If Krugman had made the standard Keynesian argument that more fiscal stimulus boosts NGDP, then he’d just need to assume sticky wages to get more hours worked.

    He is assuming wage stickiness.

    But here he seems to be making an optimal consumption smoothing argument that requires people to work more hours….

    He is also assuming optimal consumption smoothing. He has to, in order to meet the argument he believes Lucas to be making.

  47. Gravatar of Saturos Saturos
    12. December 2012 at 07:08

    Kevin, it’s not possible to increase the aggregate quantity of all GDP-goods demanded without either supplying more goods in aggregate (Say’s Law) or by increasing M*V (aggregate spending of money on GDP). If your model says otherwise then it is wrong.

  48. Gravatar of Saturos Saturos
    12. December 2012 at 07:14

    Bob, you’re right, but as I said Krugman is getting Lucas wrong, Lucas is talking about perfect crowding out of spending even without Ricardian effects. Remember the famous Cochrane essay? Why shouldn’t the multiplier be zero? Krugman responded to that elsewhere by reminding his readers that there is a glut of loanable funds at the ZLB. Scott correctly points out that Lucas is forgetting that fiscal policy has ways of raising money velocity. But Lucas is correct in thinking about where the money comes from, it’s a step out of the illogical muddle of old-Keynesian thinking.

    Scott, actually I don’t think Krugman was talking about a real mechanism. He is simply pointing out that the balanced budget multiplier (with Ricardian equivalence) is not zero.

  49. Gravatar of Kevin Donoghue Kevin Donoghue
    12. December 2012 at 09:30

    Saturos/Scott

    Yes, I was wrong to say fiscal stimulus will leave V unchanged. Obviously the change in Y generally implies a change in V. However I still disagree with this: “Krugman should have said that Lucas forgot that fiscal stimulus can impact V.” I don’t think that Lucas needed to be told that.

    “[Krugman] is simply pointing out that the balanced budget multiplier (with Ricardian equivalence) is not zero.”

    That’s it, exactly.

Leave a Reply