Two kinds of economist

I learned an important lesson by reading newspapers from 1933 showing Wall Street’s reactions to FDR’s New Deal policies.  They were sort of OK with his outrageously statist NIRA, which had the government herd companies into cartels in order to raise prices (although interestingly they weren’t too crazy about the later high wage add-on.)  But Wall Street was absolutely apoplectic about the one FDR policy that actually worked–dollar depreciation.  Tinkering with the value of the dollar–which had been fixed to gold for 54 years–was considered an outrage.  But then I noticed something interesting; every time the dollar fell a bit more after some sort of action and/or signal from the Administration, stock prices soared.  Wall Street was like some sort of masochist.  Ow! . . . hmmm, I like that, hit me again. 

This all made me think of some recent Krugman posts (and no, not for the reasons you are thinking.)  If you read Brad DeLong, you probably notice that he is a bit in awe of Krugman’s ability to be right about everything.  Actually, Krugman isn’t right about everything, but he tends to be wrong about exactly the same things that DeLong is wrong about, and so DeLong wouldn’t notice those things.  But to give the devil his due, he is right about an awful lot of things.  Why is that?  My conservative readers may assume that Krugman sold his soul to the devil, but I have a more plausible explanation.

Think about all his recent posts mocking the conservative fear that big deficits will lead to higher interest rates.  What evidence does Krugman use?  He cites the low and falling 10 year bond yields.  In other posts he has used TIPS spreads to explain why inflation is the last thing we should be worried about.  Now flash back to March 2009, when Krugman warned that $780 billion in stimulus would not be enough to get the job done.  Did he know this from his models, as he claimed?  Or did he cheat, did he peek at the equity, commodity and bond markets, and notice that all were predicting a severe recession with lots of disinflation, if not outright deflation?  I think he peeked. 

My theory is there are two kinds of economists:

1.  Those who look smarter than they really are, because they rely on the EMH to predict

          Paul Krugman

          Scott Sumner

          etc

2.  Those who look dumber than they really are because they rely on their own models to predict:

           Conservatives predicting inflation based on Quantity Theory models or Fiscal Theory models.

          etc

I bet you never thought you see a list that had Paul Krugman in the pro-EMH camp.  But just as with Wall Street in 1933, look at what people do, not what they say.  Just as I am a pro-EMH guy who occasional tries market timing in my personal investments, Krugman is an anti-EMH guy who forecasts as if he believes in the EMH.  And that makes him a very good forecaster, and very dangerous to us right-wingers.  We underestimate him at our peril.

Here is an example of a conservative who wasn’t careful.  The passage is a Greenspan quotation, cited by Krugman:

Despite the surge in federal debt to the public during the past 18 months—to $8.6 trillion from $5.5 trillion—inflation and long-term interest rates, the typical symptoms of fiscal excess, have remained remarkably subdued. This is regrettable, because it is fostering a sense of complacency that can have dire consequences. 

And here is Krugman’s reply:

You know, some people might take the fact that what’s actually happening is exactly what people like me were saying would happen — namely, that deficits in the face of a liquidity trap don’t drive up interest rates and don’t cause inflation — lends credence to the Keynesian view. But no: Greenspan KNOWS that deficits do these terrible things, and finds it “regrettable” that they aren’t actually happening.

Ouch!  Actually, deficits are bad because they lead to higher future taxes, not higher inflation.  Inflation is determined by monetary policy, but perhaps Mr. Greenspan forgot that.

BTW, it seems like half the time Krugman is pushing fiscal stimulus, and the other half of the time he is showing how countries with more stimulative policies (such as Iceland and Britain) are doing slightly less bad.   But has anyone noticed that the successful policy examples that he cites relate to monetary policy, not fiscal policy.  Just saying.


Tags:

 
 
 

58 Responses to “Two kinds of economist”

  1. Gravatar of foosion foosion
    1. July 2010 at 05:56

    EMH is one of those things that are not very good, but are better than all available alternatives.

  2. Gravatar of Paul Zrimsek Paul Zrimsek
    1. July 2010 at 06:37

    Krugman’s revealed preference for the EMH has not always been as strong as it is now.

  3. Gravatar of Doc Merlin Doc Merlin
    1. July 2010 at 07:36

    Good article. I agree about using markets to predict, we do however need better ways to figure out what markets are saying.

    I didn’t like this part though.
    “But then I noticed something interesting; every time the dollar fell a bit more after some sort of action and/or signal from the Administration, stock prices soared. Wall Street was like some sort of masochist. ”

    Naturally, stocks, being real assets and not nominal assets will go up in value during devaluation!

  4. Gravatar of Doc Merlin Doc Merlin
    1. July 2010 at 07:41

    I should explain a bit more. Devaluation tends to lower interest rates /and/ lower real value of nominal assets. This one-two combination means that real assets become more highly sought after during devaluation, both for store of value /and/ because their returns have been inflated (relative to nominal assets like debt/bonds/etc).

    This means during money_expansion/devaluation, stocks are more highly valued than they would otherwise be.

  5. Gravatar of Doc Merlin Doc Merlin
    1. July 2010 at 07:44

    For risk of gyob,

    “Actually, deficits are bad because they lead to higher future taxes, not higher inflation. Inflation is determined by monetary policy, but perhaps Mr. Greenspan forgot that.”

    Its more that higher deficit/gdp ratios are correlated with higher inflation in the future. Its disingenuous of krugman to use the “we can always inflate the debt away” argument and also use the “deficits don’t cause inflation” argument. You have to pick one, and stick with it.

  6. Gravatar of Benjamin Cole Benjamin Cole
    1. July 2010 at 09:00

    An excellent post. I am reading a econ blog right now where the author, a self-styled and smart conservative, just knows that monetary inflation is inevitable, and then he clutches at straws like gold prices to show it is coming. I think people often confuse their politics or ideals with their economics.

    The assets of America are worth less than two and even 10 years ago (property and Dow), and lenders have nowhere to lend. Unit labor costs are going down, not up. Business in America pays less today for labor and rents than two years ago, and everyone faces too much competition. How do you get inflation in this scenario?

    And even at that, I would rather take an inflationary boom right now than a prolonged recession. Fact is, property inflation would help an awful lot of lenders and investors. And yes, inflation would help pay down the national debt. There, I said it out loud.

  7. Gravatar of Benjamin Cole Benjamin Cole
    1. July 2010 at 09:00

    An excellent post. I am reading a econ blog right now where the author, a self-styled and smart conservative, just knows that monetary inflation is inevitable, and then he clutches at straws like gold prices to show it is coming. I think people often confuse their politics or ideals with their economics.

    The assets of America are worth less than two and even 10 years ago (property and Dow), and lenders have nowhere to lend. Unit labor costs are going down, not up. Business in America pays less today for labor and rents than two years ago, and everyone faces too much competition. How do you get inflation in this scenario?

    And even at that, I would rather take an inflationary boom right now than a prolonged recession. Fact is, property inflation would help an awful lot of lenders and investors. And yes, inflation would help pay down the national debt. There, I said it out loud.

  8. Gravatar of Scott Sumner Scott Sumner
    1. July 2010 at 09:02

    foosion, Certainly better than the alternatives.

    Paul, He learned his lesson from that unfortunate prediction–never bet against the bond market.

    Doc Merlin, I should have emphasized that even real stock values increased. I don’t think that interest rates fell significantly.

    There is a distinction between saying we can inflate our debts away, and saying it is likely we will have to. Is he making both claims?

  9. Gravatar of Indy Indy
    1. July 2010 at 09:09

    I’ll add my reasons to:

    “Actually, deficits are bad because…”

    1. Interest Rate Exposure Risk. If you borrow four trillion dollars and hold half of it in short-term maturing bills, *and* you entirely expect to have to roll those debts over when they come do because you’ll still be in deficit and issuing even more debt – then there is the risk that, when you do have to roll it all over, interest rates will have gone up and the cost of servicing those debts will be unexpectedly high and/or cause even higher deficits.

    Isn’t this the jam in which some of the European countries are finding themselves – having borrowed immensely at low “no worries!” rates, only to find that markets are no longer so complacent about their risks and demand enormous premiums to roll the debts over? Isn’t this a large part of the history of the IMF? Is there a good argument for how it can it never happen here (besides radical devaluation, of course)?

    If the US was about to move most of its liabilities into 2.91% ten-years, or 3.85% 30-year bonds, then I’d be less worried about this exposure, but right now we’re heavily short-term. I think we’re about 2 trillion, or 25% of the total, bills vs notes and bonds right now. That’s a lot of supply that has to be refinanced every year. Time for some extra large 10 and 30 year auctions, it seems to me. Especially if the Fed gets their act together later and makes up for lost time.

    2. Fundamental moral reasons of inter-generational integrity, fairness, and sense of legitimacy. If our grandkids end up better off because of deficits today, despite their higher taxes, because our deficit spending pays dividends to them of some sort, I’m all for it. If we’re just financing present consumption at their expense and without their consent, it’s immoral.

    Isn’t this the same argument about global warming? “Oh, but we create huge amounts of economic growth burning that carbon!” “Yes, but you may be burning unwisely, and create problems for your descendants that may be even greater than the economic growth you are delivering to them – and that is immoral, to be living high at their expense.”

    Why is one argument favored amongst conservatives and the other among liberals – and to each others exclusion? Is watching out and being careful with the mechanism (debt, fossil fuels, or what have you) by which we create current economic growth today – so that we don’t screw over our posterity tomorrow – important all the time, or isn’t it?

    I’ll add an anecdotal example that leads me to think we are not spending our “stimulus” dollars wisely at all, at least not from the point of view of our descendants. An example – here in Lawrence, Kansas, last winter’s storms has left many of the roads in a state of pot-holed and expedient-patch disrepair. And yet, there is no money this year to fix the roads. Or pay the teachers, or a host of other things.

    On the other hand, the sidewalks are just fine, but crews are busy tearing them up and pouring brand new ones. Why!? Well, the sidewalks are paid for with federal stimulus money under the “parks and recreation” segment because they are technically “trails”. Just absurd.

    There is a chance, I suppose, this will “goose” AD back onto it’s trend – but fixing the roads instead of the sidewalks would have done the same.

    The debt would have to be paid back by my kids, but they would have had to fix the roads eventually anyway – so the stream of payments is the same. But now, they’ll have to pay the debts *and* fix the roads too. I think we screwed them for a few extra man-years of current employment and consumption for our own benefit – and we probably didn’t even achieve the sustainable goose by doing so, anyway.

    Had the money been granted to the local community to spend as we chose – we all know what we would have chosen to spend it on. (This is a similar argument to the notion that the deficits should have gone into tax credits or rebates so that individuals could choose how to allocate them – though with a potential savings problems)

    We know what would actually have the most value, and it is the opposite of what we are actually doing. That is reckless, and if we can’t do better than this, we can’t justify more fiscal stimulus – even if, in a perfect world, it would have been just the thing we needed.

  10. Gravatar of Joe C Joe C
    1. July 2010 at 09:21

    So, if we are indeed, as Krugman says, in a liquidity trap, and deficits in this environment do not spur increases in interest rates and inflation. I would think ‘increasing’ deficits would actually cause deflation or at least keep the rate of inflation constant.

    This is sort of what the data, TIPS, M2 and such are displaying; does this lend credence to Krugman’s argument that we are indeed in a liquidity trap?

  11. Gravatar of Doc Merlin Doc Merlin
    1. July 2010 at 09:35

    @Benjamin Cole
    “Unit labor costs are going down, not up.”
    Untrue. Fixed labor costs per worker are going way up. Median salaries for workers (including benefits) are also up this recession. http://www.heraldnet.com/article/20100505/BLOG31/100509937

    “The assets of America are worth less than two and even 10 years ago (property and Dow), and lenders have nowhere to lend. Unit labor costs are going down, not up. Business in America pays less today for labor and rents than two years ago, and everyone faces too much competition. How do you get inflation in this scenario?”

    We do in fact have price inflation, it isn’t low either, over the last year we have had fairly average inflation as shown by CPI-U.

    “Fact is, property inflation would help an awful lot of lenders and investors. And yes, inflation would help pay down the national debt. There, I said it out loud.”

    In developmental economics we find that monetary inflation leads to excessive state borrowing in the long run. So, I am not convinced in the long run, that it would help pay down the debt.

    @Scott:
    Do you have interest rate data I could look at for that time? Or know where I could get it? I’d like to tease apart how much was people switching to real assets from nominal assets and how much was the view that the devaluation would help the industries associated with the stock.

  12. Gravatar of Ted Ted
    1. July 2010 at 09:39

    My guess is that Krugman actually doesn’t believe formally in the EMH. He probably accepts the random walk hypothesis that allows for investor expectations to matter – but he probably doesn’t accept the full blown consequences of the EMH. When Fama and company were testing the EMH they argued that a stock prices reflects all available information, what they called “efficient.” Well, to test that you needed some benchmark economic theory, so they chose CAPM. Then a bunch of researchers (including Fama himself) found that it simply didn’t work. So they either had to jettison the “price is right” version of the EMH or they had to admit that price volatility arbitrage is indeed possible. A lot of data came in and a lot of people concluded that volatility arbitrage probably wouldn’t work, and so they rejected CAPM in favor of the EMH.

    The problem is, then you have no benchmark to test whether the EMH says anything useful about prices. In fact, all of the “modern” pricing models like Fama-French 3-factor / Carhart four-factor / arbitrage pricing etc. assume that prices are right than derive a model, but they have no reason to suppose that because they rejected CAPM in favor of the EMH.

    The EMH, in its current form, empirically says basically nothing more than nobody knows what the market will do today – which is just the random walk hypothesis. There is really no way to know whether the market is aggregating information without a pricing model – and we don’t have one. Of course, I think it’s reasonable to think the market does aggregate information, but I think we have to go beyond a silly representative framework to really get at what is going on since investors invariable interpret market signals differently from one another and the market is then aggregating various different interpretations of market signals (possibly it’s averaging it out of sorts?).

    Of course, you are more careful with the way you use the EMH than most. In fact, I don’t think you’ve ever really used the EMH, in its strong-form at least (and in its weak-form, it’s actually not the EMH but everyone calls it that). You really use the random walk hypothesis and basically apply a minimal form of rational expectations to investor behavior (i.e., they respond to monetary policy). That’s fine for your purposes since you appear more interested in price movements rather than whether the actual valuation is correct.

    My view is that the EMH will probably be rejected some day. It needs a more realistic model that somehow captures the random walk hypothesis, augmented with some degree of rational expectations with heterogeneous investors who interpret market information differently and some degree of “limits of arbitrage.” But I think we are a long way off from such a realistic theory, so I guess the EMH holds as a decent approximation till then.

    You are in luck though since this problem with the EMH probably wouldn’t apply to an NGDP futures market since NGDP is relatively stable and so ‘price is right’ EMH would hopefully reasonably hold.

  13. Gravatar of Doc Merlin Doc Merlin
    1. July 2010 at 10:03

    @Joe C
    “This is sort of what the data, TIPS, M2 and such are displaying; does this lend credence to Krugman’s argument that we are indeed in a liquidity trap?”

    Monetary aggregates are somewhat fubar now because of the fed’s decision to pay interest on reserves. It is difficult to say exactly what M2 means right now because M1 and AMB are so messed up.

    In terms of CPI inflation, we have seen fairly normal CPI inflation during the last year. The deflation happened very quickly in the middle of the recession (the deflation mostly ended in late 2008), following very high PPI inflation in the early parts of the recession.
    ——————-
    Last few things:
    1. I may be a bit biased, since I don’t believe in liquidity traps of the Keynesian sort. I don’t believe in the zero lower bound: negative real rates are quite possible, and have happened on numerous occasions.

    2. I have never ever seen any evidence that expansive fiscal policy actually
    works.

    3. Also, by Keynesian theory, a liquidity trap would mean that we had zero nominal interest rates. We don’t. The overnight rates are close to zero, but all the longer term rates are well above zero.

    4. At this point we aren’t in a standard deflationary recession anymore. We are in a period of long term unemployment combined with occasional equity market trouble. You can call it a recession if you want, but it doesn’t look like what we normally call one. For example: GDP is up, and has been steadily climbing for a year. So even if I believed in Keynesian fiscal stimulus, I would say we shouldn’t have any now.

  14. Gravatar of Joe C Joe C
    1. July 2010 at 10:12

    @Doc Merlin
    “We do in fact have price inflation, it isn’t low either, over the last year we have had fairly average inflation as shown by CPI-U.”

    CPI-U may be average over the last year but has decrease the last two months, month/month change being negative and core CPI’s rate of change, on average, from the beginning of the recession, continues to decline, year/year.

    I should have included in my earlier post that inflation expectations seem to be decreasing. I don’t know the answer to this but has there been any time in our, or another country’s economic history, where there was deflation correlated with increasing deficits?

  15. Gravatar of q q
    1. July 2010 at 10:45

    krugman needs to get off this particular soapbox – it looks like he has lost the political war wrt stimulus.

    if it’s that he felt hopeless before about quantitative easing type strategies so he wouldn’t advocate them, now there are in his mind two equally hopeless options, so he is free to choose the other one.

  16. Gravatar of Benjamin Cole Benjamin Cole
    1. July 2010 at 10:52

    Doc Merlin–

    You are citing an odd unscientific regional survey from the state of Washington? Why?

    Here are the official national figs:

    Productivity and Costs

    June 03, 2010
    Productivity increased 2.8 percent in the nonfarm business sector during the first quarter of 2010 as unit labor costs fell 1.3 percent (seasonally adjusted annual rates). In manufacturing, productivity grew 1.5 percent while unit labor costs fell 1.5 percent. More…

    And on inflation, it is dead. Dead. We are heading towards flat–producer prices are soft, unit labor costs are going down, commercial rents of all kinds are below year and two-year ago levels.

    Doc, if you are looking for national data on wages, please go to the Bureau of Labor Statistics, not some goofball survey in Washington state.

  17. Gravatar of Wonks Anonymous Wonks Anonymous
    1. July 2010 at 11:44

    Will Ambrosini thinks EMH is a “subsidiary hypothesis” that allows testing of other, “joint”, hypotheses:
    http://www.ambrosini.us/wordpress/2009/10/emh-and-the-market/

    Benjamin Cole, which conservative blog are you referring to?

  18. Gravatar of Jon Jon
    1. July 2010 at 12:36

    Scott, how do distinguish stocks rising due to inflation expectations? It’s not enough to look at stock indexes alone. I think you tend instead to see comovement in stock and bond yields.

  19. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    1. July 2010 at 12:52

    It’s quite obvious that Krugman doesn’t believe in EMH and he uses his own models when he makes successful forecasts (housing bubble, problems for Bernanke in Oct. 2008, etc.)

  20. Gravatar of Benjamin Cole Benjamin Cole
    1. July 2010 at 13:15

    Wonks Anonymous–

    There is a very good econ blog, Calafia Beach. Lots of good info and insights. I think the author, erroneously, believes that inflation is inevitable due to an overly accommodative monetary policy, and this is what is spooking the market (along with all things Obama)
    Therefore Califia Beach is calling for higher interest rates etc.

  21. Gravatar of Doc Merlin Doc Merlin
    1. July 2010 at 14:58

    @Benjamin:
    1. Rereading it, I seem to have made a pretty serious mistake. For the washington study, I’ll just blame my flu as to why I wasn’t thinking straight. I misread unit labor costs as just “labor costs.” My mistake.

    Anyway from the BLS:
    “Unit labor costs fell 4.2 percent over the last four quarters, as the 6.1 percent increase in output per hour over that period outpaced a 1.6 percent rise in hourly compensation (chart 2, tables A, and 2).”
    http://www.bls.gov/news.release/pdf/prod2.pdf

    Some of the inflation hawks have a different issue. They influenced by Austrians who define “inflation” as “excessive monetary ease”, not as price index increases (the way everyone else does), and say that “excessive monetary expansion” causes bubbles.

    Anyway, you were correct wrt unit labor costs.

  22. Gravatar of Bob M. Smith Bob M. Smith
    1. July 2010 at 16:03

    Scott,

    What is the correct way to look at US borrowing costs? Sure, the 10 year treasury yield is low by historical standards, but that is a nominal rate.

    If it is a fallacy to say that the fed funds nominal rate is “low” at 0%, then isn’t a fallacy to say that the US 10 year borrowing rate is “low” at 3%?

  23. Gravatar of JimP JimP
    1. July 2010 at 17:22

    Samuel Brittan puts in a plug for nominal gdp targeting.

    If only someone at the Fed would listen.

    http://www.ft.com/cms/s/14e20968-8545-11df-9c2f-00144feabdc0,dwp_uuid=c79bdd44-33fc-11da-adae-00000e2511c8,print=yes.html#

  24. Gravatar of Doc Merlin Doc Merlin
    1. July 2010 at 17:26

    Thanks JimP. It seems we are being heard! Woot!
    And a grats to Scott, seems someone out there is listening!

  25. Gravatar of JimP JimP
    1. July 2010 at 18:18

    Martin Wolfe can, I think, also be said to agree. As Scott has said, the Brits are often more aware of our deflationary lust than we are.

    http://blogs.ft.com/martin-wolf-exchange/2010/06/27/is-monetary-policy-too-expansionary-or-not-expansionary-enough/

  26. Gravatar of Mr. E Mr. E
    1. July 2010 at 21:21

    The U.K. had extremely high debt to GDP spending. The other countries that did well had tons of cash money flowing in from commodities (Australia, Canada) – essentially this is the same as government fiscal balance expansion, right? Some magical outside force is just spending money on stuff – be it government or mysterious foreigners, it is just money appearing out of nowhere for that individual country.

    Where is the mystery?

  27. Gravatar of OGT OGT
    2. July 2010 at 04:43

    I don’t think the EMH/non-EMH distinction holds, partly for reasons Ted points out above.

    We could make a distinction between empirically inclined economists and theory inclined economists. In many ways that distinction holds up better, afterall there were a few economists that both called the housing bubble and the current disinflation.

    Rogoff and Reinhardt’s work comes to mind in that case, two economists who have poured through data over time and used that to inform ideas about about how the economy works. ‘This Time Is Different’ is not the title of an EMH book, but it is an inductive reasoning based book. No economist can be successful without some theory to help inform what the market and history is telling them, of course, but the theory should be modest before the facts.

  28. Gravatar of scott sumner scott sumner
    2. July 2010 at 05:32

    Benjamin, Yes, Krugman likes to quote Hawtrey from the Great Depression: “Crying fire, fire, in Noah’s flood.”

    Indy, Good points.

    Joe C. It depends how you define liquidity trap. If you mean that short term rates are near zero, then the answer is yes. If you mean we are “trapped” and monetary stimulus won’t work, then the answer is no.

    Doc Merlin, The Fed focuses on core inflation, and that has been running below normal for 18 months.

    A web site called “St Louis Fred” has lots of interest rate data–regular and TIPS.

    Ted, I’ve done lots of posts on the EMH. You said;

    “The EMH, in its current form, empirically says basically nothing more than nobody knows what the market will do today – which is just the random walk hypothesis.”

    I think it is quite useful The EMH says you’d do just as well in an index fund, as paying an expert to manage your investments. It says regulators won’t be able to spot bubbles. It says new information gets immediately incorporated into stock prices. It say technical analysis is bunk. I found the EMH to be very useful in my research on the Great Depression.

    I don’t think it will ever be replaced. If it was replaced with a theory of how to beat the market, that theory would breakdown almost immediately.

    You don’t need to guess that Krugman doesn’t believe in the EMH, he says so.

    q, I made that exact point a year ago. Afterwards Krugman did start mentioning monetary policy a bit more often, but still not as often as I’d like to see. Most of his readers still don’t realize that monetary stimulus is an option.

    Wonks Anonymous, Yes, I recall that Ambrosini piece. I think he is right.

    Jon, You asked;

    “Scott, how do distinguish stocks rising due to inflation expectations? It’s not enough to look at stock indexes alone. I think you tend instead to see comovement in stock and bond yields.”

    Exactly, you must always look at multiple markets to understand what is really going on. Because inflation expectations are still low, I think most of the stock rise after March 2009 was expectations of more real growth (and vice versa recently.)

    123, You said;

    “It’s quite obvious that Krugman doesn’t believe in EMH and he uses his own models when he makes successful forecasts (housing bubble, problems for Bernanke in Oct. 2008, etc.)”

    I don’t follow this at all. In October 2008 the markets were screaming that Bernanke was far too contractionary.

    Doc Merlin, The problem with Austrians is that (as far as I know) they don’t have a good definition of “monetary ease.”

    Bob, You’d want to look at real interest rates, which are now extremely low (about 1% on the 10 year, I believe.)

    Thanks JimP, Even in the Great Depression the Brits were ahead of us on monetary policy. (The real side of their economy is not doing so well recently, however.)

    Mr E, I agree that commodities helped, but recall that commodity prices fell nearly 60% in late 2008, so they also took a hit. The Aussies had a high trend rate of NGDP growth, and thus higher trend interest rates. This allowed them to avoid the zero bound.

    OGT, See my response to Ted. It is very likely that economists who called the housing bust just got lucky. When that happens, their next prediction usually won’t turn out as well (Roubini for instance, who missed the equity rally last year.)

    I think any good economists needs a mixture of theory and empirics.

  29. Gravatar of Scott Sumner: Looking for Trouble «  Modeled Behavior Scott Sumner: Looking for Trouble «  Modeled Behavior
    2. July 2010 at 08:16

    [...] ~ July 2nd, 2010 in Economics | by Niklas Blanchard Scott Sumner has a recent post about two types of economists; economists that predict with models, and economists that infer [...]

  30. Gravatar of happyjuggler0 happyjuggler0
    2. July 2010 at 09:56

    Scott,

    It is very likely that economists who called the housing bust just got lucky. When that happens, their next prediction usually won’t turn out as well (Roubini for instance, who missed the equity rally last year.)

    You need to read this article, or at least the following quotes. Roubini didn’t put his money where his mouth was. I’ll let readers chew on the implications of that.

    http://www.newsweek.com/2008/12/19/nuriel-roubini.html

    He predicted mortgage defaults would cause financial institutions to fail, and that soaring oil prices, combined with falling home prices, would cause debt-ridden consumers to dramatically cut spending[...]“[My view] was so obvious, I don’t know how anyone could argue otherwise,” he says.

    So where did he put his own money? It’s obvious isn’t it?

    Despite his prescience, he’s suffered just like the rest of us: he’s remained fully invested in stock index funds through the market downturn, causing his portfolio to plummet[...]Stocks have further to fall, he says[...]

    Ok, I guess it wasn’t obvious where he would put his money given the alleged obviousness of his views.

  31. Gravatar of Scott Sumner and the Limits of Arbitrage « Rortybomb Scott Sumner and the Limits of Arbitrage « Rortybomb
    2. July 2010 at 10:10

    [...] Sumner and the Limits of Arbitrage Posted in Uncategorized by Mike on July 2, 2010 Scott Sumner: ….Think about all his recent posts mocking the conservative fear that big deficits will lead [...]

  32. Gravatar of steve roberts steve roberts
    2. July 2010 at 11:18

    I’d say there are 2 different type of economists in a very different way:

    1. There are Keynesian economists.
    2. There are former Keynesian economists who recognize the fact Keynesian economists NEVER see a point where we shouldn’t be stimulating the economy to some extent.

  33. Gravatar of StatsGuy StatsGuy
    2. July 2010 at 11:46

    FYI, since we’re speaking of Krugman, he happened to link to an excellent example of a 20% wage cut for public sector employees that was not matched by decreases in nominal debt obligations.

    http://www.nytimes.com/2010/06/29/business/global/29austerity.html

    Unions complied – no social disorder. And as Krugman notes, ireland ain’t doing so well…

    http://krugman.blogs.nytimes.com/2010/06/30/the-icelandic-post-crisis-miracle/

    Also, Ireland started with a public debt to GDP ratio of 20%, AND low corporate taxes, AND a small fiscal surplus.

    Ireland is almost a perfect controlled experiment…

    Though, I suppose it’s possible that Ireland simply didn’t cut wages enough. (Perhaps a 40% cut in wages would have worked…)

    Anyway, yet more evidence that problem #1 is nominal debt stickiness, not nominal wage stickiness.

  34. Gravatar of Jimmy Jimmy
    2. July 2010 at 12:45

    Professor Sumner,

    You’re suffering a very severe case of confirmation bias. The stock market routinely makes predictions of recovery and recession that turn out to be wrong. This is an easily testable hypothesis, not one that we have to speculate about.

  35. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    2. July 2010 at 14:45

    Scott, you might be interested in the FT piece on NGDP by Samuel Brittan:
    http://www.ft.com/cms/s/0/14e20968-8545-11df-9c2f-00144feabdc0.html

    Unfortunately he is neutral on level targeting and neglects the role of NGDP expectations.

  36. Gravatar of Lorenzo from Oz Lorenzo from Oz
    2. July 2010 at 14:56

    Jimmy: The stock market routinely makes predictions of recovery and recession that turn out to be wrong. This is an easily testable hypothesis, not one that we have to speculate about.
    The question is not “what is a perfect predictor?”, there is no such thing. The question is, “what uses available information better?”. Scott’s point is that market information continues to be better than the output of models.

    See also
    this critique of models in general. Which is not the same as arguing that all models are false. I have a long post about problems with (computer) models here.

  37. Gravatar of scott sumner scott sumner
    2. July 2010 at 18:31

    Happyjuggler), Yes, I also heard that story. I find it quite revealing.

    Steve, That was more of a problem when I was younger. I think the smarter Keynesians no longer favor stimulus during periods of low unemployment.

    Statsguy, You said;

    Ireland is almost a perfect controlled experiment…

    Though, I suppose it’s possible that Ireland simply didn’t cut wages enough. (Perhaps a 40% cut in wages would have worked…)

    Anyway, yet more evidence that problem #1 is nominal debt stickiness, not nominal wage stickiness.”

    Actually, it isn’t evidence of anything, and for several different reasons:

    1. The sticky wage transmission mechanism refers to private sector wages, not public sector wages. The reason is as follows. Wage cuts lower the MC curve for a given firm. If the wage cuts don’t lower AD, then output will increase. But governments don’t produce where MC=MR, hence wage cuts have no necessary effect on government output.

    2. Unless the government is revenue constrained, then it is almost a tautology that wage cuts increase output. Suppose the Irish government has $500,000,000 to spend on teachers salaries, and that is all. Then if they cut teacher wages from $50,000 to $25,000, they can double teacher employment from 10,000 to 20,000, at the same level of government spending.

    3. Now you might argue that the cut in salaries would be combined with a cut in total government spending, but that would not be any sort of controlled experiment, as you’d be combining less wage stickiness and an adverse AD shock at the same time.

    4. Sticky wages don’t cause all recessions, they cause recessions when there are nominal shocks. Ireland isn’t like the US. It’s about the size of metro Detroit or metro Phoenix. And just like those two cities, it would have had a recession even if wages were perfectly flexible. It was hit hard by a real shock.

    5. I still think Mexico’s 99.9% reduction in NGDP in 1993 is the best experiment showing the important of wage and price flexibility. It’s true that nominal debts were also adjusted, but most of the recessions in my lifetime have not featured debt crises, so I don’t see that as a plausible transmission mechanism for nominal shocks.

    6. I don’t think the government should be trying to get the US out of recessions by cutting wages, they should be trying to boost NGDP by an expansionary monetary policy. Ireland has a difficult choice to make–do they want to stick with the euro. Only if the answer is yes do wages need to be cut.

    7. If they stay in the euro then they have no choice but austerity. Otherwise they face bankruptcy. But if they are going to declare bankruptcy, then they really should be leaving the euro instead.

    Jimmy, Nobody is very good at predicting recessions, but the stock market is no worse than anyone else. If recessions were forecastable, then they would not happen.

  38. Gravatar of scott sumner scott sumner
    2. July 2010 at 18:38

    123, Thanks, several people have sent that one. I’ll try to link to it soon.

    Lorenzo. I agree with your post. It looks like you have a very good blog.

  39. Gravatar of Benjamin Cole Benjamin Cole
    2. July 2010 at 20:05

    Doc Merlin-

    Hey, I made mistakes worse than that. Onward qualitative easers!

  40. Gravatar of Doc Merlin Doc Merlin
    2. July 2010 at 23:33

    @Benjamin
    I’m not really a qualitative easer, I think its far too late for monetary expansion to cause short term help at this point. I definitely support NGDP targeting through a market mechanism, however, and think it is a far better policy than what we currently do.

    Also (this may be a pipe dream), I’d like to get rid of the fed altogether and have free market in currency, as that (theoretically) approximates NGDP targeting with a market mechanism, without the problems that having a monopoly entails.

    This makes me an disestablishmentsumnerian, I guess.

  41. Gravatar of Doc Merlin Doc Merlin
    2. July 2010 at 23:36

    Re: Scott,
    That is a very good point about government pay from a microeconomic angle. I need to copy that down!

  42. Gravatar of Left Outside Left Outside
    3. July 2010 at 03:34

    “disestablishmentsumnerian”

    That is pure genius.

  43. Gravatar of More Post-Modern Financial Analysis From Scott Sumner More Post-Modern Financial Analysis From Scott Sumner
    3. July 2010 at 06:53

    [...] Here’s Scott on those nutty goldbugs: [...]

  44. Gravatar of ssumner ssumner
    3. July 2010 at 07:14

    And those who disagree with Doc and I are antidisestablishsumnerians and their philosophy is therefore antidisestablishsumnerianism, which is tied for the longest word in the English language (and perhaps the most useless as well.)

  45. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    3. July 2010 at 07:58

    “I don’t follow this at all. In October 2008 the markets were screaming that Bernanke was far too contractionary.”

    Krugman has explicitly said he thinks EMH is harmful. Of course this doesn’t mean that he thinks that market prices have no informational content, he just thinks that markets reflect information inefficiently to some degree.

    On September 17 2008 Krugman was concerned with the macro consequences of Lehman blowup. He noted that messages S&P 500 and TED spread were sending are contradictionary, and he was more worried about the economy than the stockmarket was at that time. Krugman’s knowledge let him process the knowledge embedded in eurodollar and treasury markets better than stockmarket.

    In October things are more complicated as divergence between TED spread and S&P has narrowed somewhat, but I think that Krugman was more bearish in October than S&P 500 was.

    I haven’t read his recent op-ed called “The third depression” yet, but I think it’s clear that stockmarket isn’t signalling the third depression yet.

  46. Gravatar of ssumner ssumner
    3. July 2010 at 08:18

    123, Krugman greatly underestimated the severity of the recession in late 2008, so I’m not convinced he was all that much more bearish than investors. Stocks were down substantially by that time. But I agree with your broader point that he doesn’t alway predict in exactly the way markets are predicting. I certainly don’t agree that he forecasts better than markets do.

    Krugman has a fairly elastic definition of ‘depression.’ He includes the Japanese case, and the 1870s-1890s, both cases that most people would not define as depressions. My hunch is that he actually doesn’t expect a full-fledged depression, but is trying to go out on a limb so that if things get worse than people expected, he can say I told you so.

  47. Gravatar of StatsGuy StatsGuy
    3. July 2010 at 16:35

    ssumner:

    Points well taken, however some notes:

    “but most of the recessions in my lifetime have not featured debt crises”

    True, which is why this recession is different. Notably, most of the recessions in our lifetime have not involved the Fed deploying a ZIRP and failing to restart solid growth either, or outright deflation. My concern is whether wage decreases alone – even if very aggressive – are enough to recover from a debt crisis driven depression. Thus the importance of the transmission mechanism in _this_ recession.

    “Unless the government is revenue constrained, then it is almost a tautology that wage cuts increase output.”

    That assumes the government is optimizing based on a budget constraint, not economizing based on a required output schedule. The same is true in the private sector; note also that most labor supply models assume that increasing labor allocated immediately increases production. In complex work environments, this is often quite the opposite –

    http://en.wikipedia.org/wiki/Brooks's_law

    One wonders what would happen if the labor output curve in non-widget industries (now most of the economy) reflected this?

    Also, the Irish govt. didn’t appear to increase output.

    “cut in total government spending, but that would not be any sort of controlled experiment, as you’d be combining less wage stickiness and an adverse AD shock ”

    Isn’t that a Keynesian argument? :)

    Re: exports – without a doubt, exports are a much larger part of the Irish economy (something like 100 billion vs. 180 billion total economy), but it’s noteworthy that the Irish growth in Q1 of 2010 was led by exports to UK and US (as the Euro fell). Moreover, on the issue of whether the Irish recession was real or nominal, I submit two data points:

    - from sept to sept, 2008 to 2009, price levels dropped 6.5%.

    - The Irish experienced a large property value collapse (I won’t call it bubble, since I’m not sure what a bubble is anymore), and a banking crisis much like the US, and 12% of employees worked directly in construction, with another ~12% working in supporting sectors. So I would argue that Ireland’s crisis is indeed very nominal – at least as nominal as, say, Iceland’s.

    In any case, I agree with you on points 5, 6, 7, as you know.

  48. Gravatar of Greg Ransom Greg Ransom
    3. July 2010 at 22:04

    This is a false choice, isn’t it?

    “I’ve had an ongoing debate with commenters about whether the financial crisis was an exogenous shock caused by bad lending practices, or an endogenous response to tight money and falling NGDP.”

    The exogenous “shock” led to a collapse in “shadow money” (e.g. assets used as substitutes for money) .

    The exogenous shock was an inevitable bust of an artificial boom (as you have said happened, remember?)

    The tight money and falling NGDP in the first instance where unavoidable consequences of the unavoidable collapse of the artificial malinvestment boom.

    Things may have been made worse by bad Fed policy — but there was no avoiding the “correction” of the malinvestment boom, as you have elsewhere admitted.

    False choices do not advance understanding.

  49. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    4. July 2010 at 15:57

    “Krugman has a fairly elastic definition of ‘depression.’ He includes the Japanese case, and the 1870s-1890s, both cases that most people would not define as depressions. My hunch is that he actually doesn’t expect a full-fledged depression, but is trying to go out on a limb so that if things get worse than people expected, he can say I told you so.”

    I think he is worried about the repeat of Japanese lost decade. But stocks are overvalued if the risk of the Japanese scenario is as high as Krugman thinks.

  50. Gravatar of ssumner ssumner
    4. July 2010 at 17:35

    Statsguy; You said;

    “True, which is why this recession is different. Notably, most of the recessions in our lifetime have not involved the Fed deploying a ZIRP and failing to restart solid growth either, or outright deflation. My concern is whether wage decreases alone – even if very aggressive – are enough to recover from a debt crisis driven depression. Thus the importance of the transmission mechanism in _this_ recession.”

    Again, I think you assume that we disagree more than we do. Just to be clear, I do not think wage cutting is a good way out of this recession. Some conservatives think wage cutting is a good way out of recessions, but I am not one of them. But if government attempts to prop up wages, I do think that will make unemployment higher.

    You said;

    “That assumes the government is optimizing based on a budget constraint, not economizing based on a required output schedule. The same is true in the private sector; note also that most labor supply models assume that increasing labor allocated immediately increases production. In complex work environments, this is often quite the opposite –”

    Agreed, but that’s why I mentioned teachers. My daughter once had four teachers in a class of about 18 kids (public school.) Education has almost an infinite ability to expand jobs, just lower the student/teacher ratio.

    Regarding your Keynesian smiley face, I should have known you’d catch that one. But think of it this way—Keynesians also believe in the sticky wage transmission mechanism. So even if you are right, (and there are other objections I mentioned) then at best it shows either wage stickiness is not a problem or the Keynesian theory of fiscal stimulus is right. But it isn’t a “controlled” experiment.

    I don’t quite follow the last part of your comment, as I didn’t make any reference to exports. I assumed the real estate shock in Ireland was much bigger than in the US, more comparable to a city like Phoenix, which was one of the worst hit cities in the US. So I don’t see why you call that a nominal shock. Ireland is part of the euro region, nominal shocks are euro-wide shocks. If the price of Ireland real estate falls sharply relative to eurozone real estate, then that is a real shock and leads to some unemployment due to reallocation out of that sector. That is where people like Kling are correct, those workers can’t immediately find jobs in other sectors. I suppose Eastern European workers left the country altogether, so I’d guess the fall in GDP was bigger than the rise in unemployment.

    Greg, Well then explain why so many Austrian commenters disagree with me when I present exactly that nuanced view of things. I’ve said from day one that the original 2007-08 crisis was real misallocation, not a nominal shock.

    123, Actually, I am also sort of worried about the Japanese scenario, although for two reasons I don’t think it will be quite as bad here:

    1. We have an extra 1% population growth
    2. We have a 2% higher inflation target.

    That means we have an extra 3% NGDP growth, if not more. So we aren’t likely to be stuck in a liquidity trap for 16 years (I hope.)

  51. Gravatar of OGT OGT
    5. July 2010 at 08:43

    Sumner: “I think any good economists needs a mixture of theory and empirics.”

    True, just as any good wagon needs both a cart and a horse. It is generally also important to know which one comes first.

    That’s why economics has arguably been a value destroying industry over the last thirty years.

    Rogoff and Reinhardt:

    “The mainstream of academic research in macroeconomics puts theoretical coherence and elegance first, and investigating the data second,” says Mr. Rogoff. For that reason, he says, much of the profession’s celebrated work “was not terribly useful in either predicting the financial crisis, or in assessing how it would it play out once it happened.”

    “People almost pride themselves on not paying attention to current events,” he says.

    In the past, other economists often took the same empirical approach as the Reinhart-Rogoff team. But this approach fell into disfavor over the last few decades as economists glorified financial papers that were theory-rich and data-poor.

    Much of that theory-driven work, critics say, is built on the same disassembled and reassembled sets of data points — generally from just the last 25 years or so and from the same handful of rich countries — that quants have whisked into ever more dazzling and complicated mathematical formations…

    “There is so much inbredness in this profession,” says Ms. Reinhart. “They all read the same sources. They all use the same data sets. They all talk to the same people. There is endless extrapolation on extrapolation on extrapolation, and for years that is what has been rewarded.”

    http://www.nytimes.com/2010/07/04/business/economy/04econ.html?ref=business

  52. Gravatar of Doc Merlin Doc Merlin
    5. July 2010 at 09:11

    ‘That means we have an extra 3% NGDP growth, if not more. So we aren’t likely to be stuck in a liquidity trap for 16 years (I hope.)’

    You can look at the Japanese lost decade as a supply side problem too (ass opposed to a liquidity trap). They do after all have the highest corporate tax rate in the world, and have a system set up to protect businesses from their own failures.

  53. Gravatar of ssumner ssumner
    5. July 2010 at 11:29

    OGT, Yes, I have many of the same criticisms of mainstream economics.

    Doc Merlin, Yes, there were both supply and demand-side problems. Some of the supply-side problems resulted from a lack of AD, just as in the US.

  54. Gravatar of Greg Ransom Greg Ransom
    5. July 2010 at 23:06

    Well, I see a split among “Austrians”, with the majority lining up with you. And those who are more strongly “Hayekian” on theory line up with you close to 100% on those, and most who don’t are “Misesians”.

    Scott writes,

    “Greg, Well then explain why so many Austrian commenters disagree with me when I present exactly that nuanced view of things. I’ve said from day one that the original 2007-08 crisis was real misallocation, not a nominal shock.”

    Note also the some difference are not differences over theory, they are differences on what the facts actually are. “Austrians” are not agreed on what happened in 2008.

    And the set of facts keeps expanding.

    Fannie and Freddie are now at $390 billion in taxpayer bailouts, for example.

    We keep learning more about what happened in the past as history unfolds.

  55. Gravatar of ssumner ssumner
    6. July 2010 at 06:52

    Greg, Fine, but that doens’t contradict anything I said. I merely indicated that I had debated commenters, not that they all had the same view of Austrian economics.

  56. Gravatar of What the market wants – Economics - What the market wants - Economics -
    13. July 2010 at 13:37

    [...] something is being missed in this conversation. A little while ago, Scott Sumner caused a furore by suggesting that Paul Krugman, like himself, seemed to get a lot of things right because he relied on the [...]

  57. Gravatar of Trust markets, sometimes – Economics - Trust markets, sometimes - Economics -
    26. October 2010 at 08:32

    [...] Mr Krugman doesn't buy it. And upon reading his scepticism I was immediately reminded of Scott Sumner's point that when he and Mr Krugman get their forecasts right, it's because they trust what markets are [...]

  58. Gravatar of Trust markets, sometimes [The Economist] | DreamInn Trust markets, sometimes [The Economist] | DreamInn
    26. October 2010 at 09:26

    [...] Mr Krugman doesn’t buy it. And upon reading his scepticism I was immediately reminded of Scott Sumner’s point that when he and Mr Krugman get their forecasts right, it’s because they trust what markets [...]

Leave a Reply