Second reply to David Andolfatto

David Andolfatto has a reply to my recent post.

The BLS data show housing prices rising by 10% after housing prices peaked? Not sure I understand this claim. I thought that the price of housing services entered into the CPI, not house prices directly. In any case, it would have been nice to have been provided with an alternative price index.

Yes, the BLS measures housing prices by looking at the cost of housing services, not the price of houses.  That was my point.  Why should the price of “housing services” be put in a model that focuses on issues such as employment and output, rather than the price of new houses.  That makes no sense to me.  Labor is used in building new houses.  Very little labor is involved in housing services from older houses.  Which house price index better explains the path of housing construction over the past 10 years?

After I wrote:

But even if the data were accurate, prices are the wrong variable, and models that suggest PLT is equivalent to NGDPLT are simply wrong. Indeed one of the strongest arguments for NGDPLT is that it does better when productivity growth is unstable. And productivity growth in America is unstable.

David responded:

Scott, I hate to break this to you but: all models are wrong in the sense that they are abstract representations of reality. Perhaps you mean “wrong” in the sense that any model that displays such an equivalence necessarily does not fit the data? If so, what evidence do you have that supports this claim?

To put this in context, David had argued that some models showed that PLT is almost indistinguishable from NGDPLT.  He went on to argue that this was especially true if productivity growth was stable.  I agree that there are scenarios where the two policies are almost identical.  But productivity growth in the US is not stable.  Whenever there is a shift in the AS curve, a policy of PLT will yield a different result from a policy of NGDPLT.  In other words the burden of proof is on Andolfatto.  One can come up with all sorts of models.  Tell me why I should be interested in models that predict almost identical effects from PLT and NGDPLT.  If NGDPLT is advocated as a policy that adapts better to shifts in productivity (and it is), why should I reject it on the grounds that it isn’t much different from PLT in models where productivity is stable?

By the way, Miles Kimball, who has some kind words to offer your crowd, claims here that the NGDP target has to be adjusted for changes in productivity growth. But maybe you have some different model in mind? Where does this model live?

Yes I do, and the model lives here.  (Without the annoying math.  I don’t think mathematical macro models are useful, I gather David doesn’t agree.)  Kimball simply asserts that we should want to adjust for productivity changes in order to stabilize inflation.  But NGDPLT proponents argue that the whole point of the policy is to destabilize inflation.  A stable inflation rate is bad for the welfare of the economy.  First because it leads to greater instability in output, and second, because even the welfare costs of inflation itself are greater with a stable inflation rate than a stable NGDP growth rate.  Kimball doesn’t seem aware of these arguments, so he assumes that market monetarists simply made a mistake.

I am not sure why a call for “fiscal stimulus” in late 2008 and early 2009 would have been “madness.” The PCE price-level peaked in July 2008 and fell sharply in late 2008 and early 2009 (largely reflecting the collapse in energy prices).

That’s a defensible view, but I still tend to doubt that policy was on target from the Fed’s perspective.  I can certain point to statements like NY Fed President Dudley’s recent remark that in retrospect Fed policy should have been easier in 2009.  If Andolfatto was correct, I can’t imagine why the Fed would not be pleased with the path of AD in 2009.  I’d guess that if Andolfatto visited the Fed and claimed that nominal spending and prices were right were the Fed wanted them in 2009-13, he would meet a lot of resistance.  But obviously that’s a judgment call on my part.  Some might agree with him.

Andolfatto then shows that changing the trend line in the way I suggested made almost no difference.  My mistake.  But I would still insist that the trend line is very unreliable.  Obviously one could draw other trend lines that are reasonably good fits, just not quite as good as Andolfatto’s, and yet would lead to radically different conclusions.  To avoid a wild goose chase, with everyone wasting lots of time, just consider a trend line running through the initial point and the September 2008 level of prices.  That trend line would have an average inflation rate of 2.3% instead of 2.09%, and would be well above the actual price level since 2008.  I’m not saying that’s the “right one”, it probably isn’t.  But when the Fed met in September 2008 they decided not to change policy, and left rates at 2%.  They cited equal risks of inflation and recession.  That’s all consistent with my view that the Fed did not want lower AD as of September 2008.  So it’s not obviously less reflective of Fed views than the trend line drawn by David.

A counterargument is that 2008 prices were a bubble, with high gasoline prices.  But gasoline prices are nearly back up to September 2008 levels, and it no longer looks like a bubble.  In retrospect it is the very low gasoline prices of 2009, the low overall price level of 2009 that looks the like outlier.  Call it a “negative bubble” if you wish.  Two dollar gas and an S&P500 at 700.  That’s another way of saying that core inflation has run below 2% over the past 5 years.

I said:

If PLT and NGDPLT really were similar policies, then why does NGDP look far below trend since 2008, while the price level (according to Andolfatto, but I have my doubts) is right on trend?

And Andolfatto responded:

That would be because the RGDP is below trend. And there are many reasons why RGDP may be below trend that are independent of the conduct of monetary policy.

I tend to forget that not everyone takes it as a given that low levels of AD have depressed output since 2008.  So his point is well taken. Of course I could cite lots of people, including the Miles Kimbell post that David cited, as making strong arguments that inadequate demand has depressed RGDP in recent years.

On the other hand, let’s suppose we did NGDP targeting instead of PLT.  And let’s suppose that RGDP has been depressed for reasons having nothing to do with monetary policy.  Obviously NGDP growth in recent years would have been much higher under NGDPLT than PLT, which suggests (by assumption) that the price level would also have been quite a bit higher.  At a minimum that casts doubt on the hypothesis that PLT and NGDP would have produced similar results.

One final point.  In his first post David makes the following claim:

Seems to me that they are just asking for more price inflation and wishfully hoping that some of the subsequent rise in NGDP will take the form of real income.

There might be a sense in which this is true, but I think this it is misleading in two ways:

1.  I was an inflation hawk in the 1970s, and have never changed my views.  Conditions have changed.  Most of the other market monetarists have similar backgrounds.  Most of us are not Keynesians looking for a cloak to hide our secret inflationist tendencies.  We sincerely favor NGDPLT, and not higher inflation.  Personally, I don’t think inflation even matters.  All the so-called welfare costs of inflation are more closely correlated with NGDP growth anyway.

2.  The Fed denies it is engaged in PLT, and they certainly do not publically adhere to the target path hypothesized by Andolfatto.  Instead they inflation target using the PCE.  Right now PCE inflation is running 1.4%, or 1.2% for the more reliable core PCE.  That’s well below the Fed’s target, and they have no strategy to bring inflation back up to target.  They are about to tighten monetary policy despite forecasts that they will continue to fall short of their inflation and employment objectives.  One could just as well argue that those who do not favor easier money are “disinflationists” who oppose the Fed’s policy of “stable prices,” defined at 2% PCE inflation.

As far as my “hope” for more real income, I’d like to point out that when unemployment was 10% in 2009 I claimed it was far above the natural rate.  That was obviously an implied prediction that unemployment would fall with even halfway decent NGDP growth.  I might have been wrong.  The French unemployment rate really did stick at 10% after the two oil shock recessions.  Obama could have made us like France, and I would have been wrong.  But I was right, and unemployment is now down to 7.4%.  That’s not to say Obama has not moved us 10% or 20% in the direction of France, and raised our natural rate of unemployment by 0.5% or 1.0%.  I think he has.  But surely it’s now pretty clear that much of the problem in 2009 was inadequate AD.

And as always, if we had the Fed create and subsidize trading in RGDP and NGDP futures markets, which it could do at trivial cost, we would already know the answer to these important questions.  We could look at the real time response of expectations to monetary policy surprises.  But I’m the member of a brain dead profession, so we don’t have those markets.

(Of course the brain dead remark is not aimed at David, but rather those economists who oppose low cost policies that would yield valuable market signals.)

Update:  Michael Darda to the rescue, with a very plausible trend line that runs through the cyclical peaks of 1990 and 2007:




42 Responses to “Second reply to David Andolfatto”

  1. Gravatar of David Andolfatto David Andolfatto
    4. September 2013 at 11:15

    Scott, that trend decomposition provided by Darda is embarrassing. Looks like he just picked his start point. And more important, that blip in 2008 was entirely energy prices. Nobody expected that to be permanent. If you take the blip out, then the actual inflation path runs underneath the blue trend line since 1998.

  2. Gravatar of Daniel R. Grayson Daniel R. Grayson
    4. September 2013 at 11:40

    A good way to display the difference between a trend line y = mx + b and data y = f(x) is to plot y = 0 and y = f(x) – (mx+b) instead. Enlarging the scale vertically to fill the space sometimes results in dramatic views.

  3. Gravatar of jknarr jknarr
    4. September 2013 at 11:48

    It does not look like a regression or pick-your-points trend line — it looks like a (hypothetical) annual 2.24% inflation level (logged, cumulative ) since 1990 (same base).

  4. Gravatar of Don Geddis Don Geddis
    4. September 2013 at 11:55

    David A. wrote: “that blip in 2008 was entirely energy prices. Nobody expected that to be permanent.

    The Peak Oil people sure expected high energy prices to be the new normal.

  5. Gravatar of Kevin Dick Kevin Dick
    4. September 2013 at 11:59

    +1 to Don Geddis. In hindsight, it’s easy to tell blips from regime shifts.

  6. Gravatar of Morgan Warstler Morgan Warstler
    4. September 2013 at 12:20

    his dates seem weird. missing zeros on months?

    Either way:

    That thing does nothing to capture the overheating of 99-2000.

    Unemployment hit 30 yr lows, WHILE BECOMING MORE PRODUCTIVE and sprang back over 6% in short time.

    See here with 2000 employment levels:

    I don’t know how you draw that line as he does…. it doesn’t capture 2000 as a meaningful event.

    This graph makes 90-98 basically the flip side of 1998-2006.

    And that’s just outright contra my memory of the economy, 92-98 should be very nearer the trend line (what we should expect), coming off the recession in 91 that toppled Bush.

    And then 98 we ought to be like WTF, this makes 1/1/1998 look like nothing interesting happened.

    Shouldn’t the trend line capture that stuff?

  7. Gravatar of Pedro Pedro
    4. September 2013 at 14:49

    David is exactly right on this one, that trend line makes absolutely no sense. Even if you truncate the data to exclude post-peak observations and run a simple regression to estimate the coefficient of a straight line, you get a slope (quarterly data) of 0.00503. Darda seems to be simply rotating the trend to start at the origin and using a larger value for the slope (why?), so I can’t really make head or tails of that.

  8. Gravatar of jknarr jknarr
    4. September 2013 at 15:51

    “Nominal GDP targeting operates by stabilizing the debt to-
    GDP ratio. With financial contracts specifying liabilities fixed in terms of money, a policy that
    stabilizes the monetary value of real incomes ensures that borrowers are not forced to bear too much
    of the aggregate risk, converting nominal debt into real equity.”

  9. Gravatar of ssumner ssumner
    4. September 2013 at 16:13

    David, You said;

    “Scott, that trend decomposition provided by Darda is embarrassing. Looks like he just picked his start point. And more important, that blip in 2008 was entirely energy prices. Nobody expected that to be permanent.”

    I see nothing embarrassing at all, he’s just offering another possible target path, not saying it’s the best fit. Is your claim that the Fed decided on a 2.09% target path for PCE in 1990, and stuck with it all these years? Why 2.09%? Why not 2%. or 2.2%. What made the Fed pick 2.09%. Does the number 2.09% appear in the Fed minutes for 1990? Obviously not. So it was luck. But isn’t your claim that it was intentional? If the 2.09% was not intentional, then the 2.24% is every bit as plausible. More likely, it the policy goal was time varying.

    Does “nobody” include future markets circa September 2008? Was was the forward price of oil in September 2008? What was the spot price? How does the current price compare?

    That price of oil was based on the China boom and (at the time) no knowledge of the giant global recession on the way. You are saying it was a bubble? And yet we’ve had a catastrophe in Europe that no one expected, and still gasoline is back over $3.50 a gallon. Where would oil be today if the US and European and Japanese economies had not tanked?

    I would add that your comment is wrong for another reason. He ran his trend line through 2007, when oil prices were about the same as today, not 2008.

    Pedro, Neither you nor David seems to understand what Michael was doing. He wasn’t trying to provide the best fit. He was suggesting an alternative target path for the PCE. You are criticizing him for something that he wasn’t even trying to do.

  10. Gravatar of Noah Smith Noah Smith
    4. September 2013 at 16:18

    Scott, haven’t you argued in the past that the absence of mini-recessions in the U.S. means that the impact of productivity shocks on our economy isn’t large?

  11. Gravatar of Steve Roth Steve Roth
    4. September 2013 at 16:54

    I pulled a picture of this disparity (CPI vs Case Shiller), and did a quick look at what CPI might have looked like if the C-S index was used for the housing portion:

    Sumner: Has CPI Been Wildly Overstating Inflation?

  12. Gravatar of Morgan Warstler Morgan Warstler
    4. September 2013 at 17:10

    I don’t think anyone can argue the Housing Crisis didn’t CAUSE the Fin Crisis / Recession after you look at this graph:

    That is clearly proof that no matter when housing market “peaked” in 2006….

    Obviously the Fin Crisis recession went down bc the mortgage payments (from junk originated in 2004-2006) stopped coming in.

    The delinquency rate on even Prime ARM had doubled by then.

    So AGAIN…

    Stop saying the MM could have stepped in Fall 2008 and kept NGDP on target.

    INSTEAD SAY, if we had NGDPLT in place the housing peak of 2006 would have crested back in late 2004, as huge numbers of sub prime don’t get written in 2005 and 2006.

    Forget args over home sales. Those delinquent mortgages wouldn’t exist bc they never would have been created.

    Those delinquency rates (and the trillions of derivatives built upon that not happening) are charging at Q3 2008 at fevered pitch.

    IF we made this arg, this stuff from Andolfatto wouldn’t even matter.

  13. Gravatar of Mark A. Sadowski Mark A. Sadowski
    4. September 2013 at 17:43

    David Andolfatto, up until this point, seems to have been focusing on debt as a point of comparison between PLT and NGDPLT. PLT outperformed NGDPLT in his OLG model.

    However, Lars Svensson just released a paper showing that tightening monetary policy increases leverage (at least temporarily):

    “Leaning Against the Wind” Leads to a Higher (Not Lower) Household Debt-to-GDP Ratio
    Lars E.O. Svensson
    August 2013

    “”Leaning against the wind” “” a tighter monetary policy than necessary for stabilizing inflation around the inflation target and unemployment around a long-run sustainable rate “” has been justified as a way of reducing household indebtedness. But it actually has the opposite effect; it leads to higher real household debt and a higher household debt-to-GDP ratio. The reason is that a tighter policy than a baseline induces a very slow fall relative to the baseline of total nominal (mortgage) debt but a faster fall in the nominal price level and nominal GDP. There is then first a rise in real debt and the debt-to-GDP ratio relative to the baseline, a rise that is almost as large and as fast as the fall in the price level and nominal GDP. Then, real debt and the debt-to-GDP ratio very slowly fall back to the baseline, something that could take more than a decade. Therefore, “leaning against the wind” as a way of reducing the household debt-to-GDP ratio is counterproductive.”

    It’s not that difficult to imagine a scenario where an economy is hit by the third largest negative AS shock since WW II:

    And the central bank passively tightens monetary policy in order to stabilize the price level causing leverage to rise high enough to cause a financial crisis.

  14. Gravatar of Don Geddis Don Geddis
    4. September 2013 at 17:45

    Morgan: Housing could have crashed (or even housing and the financial sector) … but that doesn’t mean that the whole economy needed to fall into recession. The US has a huge and diversified economy. The dot-com crash in 2000 was localized to that one sector, and the stock market crash in 1987 also had no significant economy-wide effects. The housing/finance crisis in 2007/8 also COULD have remained local to those industries (if only NGDP had stayed on trend).

    Yes, I know you want to make a different argument. But you shouldn’t get there, by underestimating the power of monetary policy.

  15. Gravatar of Benjamin Cole Benjamin Cole
    4. September 2013 at 17:49

    You know, talking to economists these days about inflation and the economy is like talking to teenage boys about sex and dating.

    Inflation always comes first, sex always comes first.

    But there are these other avenues to explore, like real output.

    “Seems to me that they are just asking for more price inflation and wishfully hoping that some of the subsequent rise in NGDP will take the form of real income.” David A.

    This has it backwards. I want a big ramp-up in NGDP, as I hope it will lead to a big ramp-up in RGDP. Inflation is a side-effect, and as long as it is in single digits i don’t care that much (in this context; that is we trying to get out of the worst recession since the Great Depression).

    BTW. the 1970s was a horrible decade, right?

    Here are the real–real!–GDP growth percentage rates for that decade, starting in 1970.


    Egads! The size of the real GDP grew by 20 percent in just the last four years of the 1970s. (See table b-5 Economic Report of the President).

    Would anyone like a 20 percent increase in the GDP in the next four years?

    And since 2008, the real growth rates…


    Inflation dead since 2008, btw.

    Somehow, with all the structural imperfections of the 1970s–unionized workforces, less imports, regulated banking (anyone remember Reg Q?), regulated transportation, regulated telecommunications, a top federal income tax rate of 70 percent—the economy was able to post 5 percent real growth in many years and 20 percent real growth in just the last four years of the decade.

    Such a terrible economy back then. Oh, horrors.

    You think, with inflation dead now, should the Fed set real growth targets targets a lot higher than it is?

    Or, is “real growth” no longer a topic of concern in the economic profession?

    As containing inflation is so much more important?

    The entire economics profession needs an enema.

  16. Gravatar of Morgan Warstler Morgan Warstler
    4. September 2013 at 18:18


    Again, the Fin Crisis was / is different from 2000:

    1. Trillions of derivatives all relied on that mortgage income stream.

    2. The Fed stops thinking about inflation OR unemployment when the banks, not the shitty little business loan banks, the BIG HAIRY BANKS are under mortal threat.


    See, I’m not arguing MM is too puny to pull off a deadlift from Fall 2008.

    I’m arguing that FIRST SAYING to opponents that NDGP would have never let those mortgages get originated, is:

    1. True. It’s 100% TRUE.

    2. Keeps them from drawing lines on graphs, bc and I know this takes some thing… It’s EASIER and more powerful to prove Fin Crisis averted before its seeds even get planted.

    This doesn’t mean MM can’t argue it can handle 2008…

    It means that NOT claiming to solve the underlying real problem in 2004-6 (hell that’d be EASY in monetary terms), makes the bigger bolder claim suspicious.

    So again, SURE, NGDP could have kept nominal income going even as NO BANKS made loans… and it wouldn’t have hit everyone quite as hard.

    But, look man housing ain’t tech. It’s the biggest check every person writes each month. Scott himself screams how big it is to complain about CPI.

    Whats on people’s minds is housing crisis, and we either have a real answer that says MM would have stopped this shit, or MM is fundamentally flawed.

    MM is 49% about what to do about recessions IF they happen, its 51% about how to make sure recessions are far less frequent, and far less deep.

    Its 51% bc thats what norms want to be promised. MM is insurance against nightmares.

    And not selling it that way, instead of describing how to fight off Freddy Kruger is horrible marketing.

  17. Gravatar of More Trend ‘Hustle’ | Historinhas More Trend ‘Hustle’ | Historinhas
    4. September 2013 at 19:49

    […] the comments of Scott´s second reply to David […]

  18. Gravatar of lxdr1f7 lxdr1f7
    4. September 2013 at 20:06

    “A stable inflation rate is bad for the welfare of the economy. First because it leads to greater instability in output, and second, because even the welfare costs of inflation itself are greater with a stable inflation rate than a stable NGDP growth rate.”

    This seems completely counterintuitive. Is there any evidence for this? What is the reasoning behind this?

  19. Gravatar of Prakash Prakash
    4. September 2013 at 21:34


    Quoting from George Selgin’s “less than zero”

    What happens in the case just described if the authorities, instead of stabilising spending, attempt to stabilise the price level? Then, rather than let the economy come to rest at its natural’ equilibrium, d(n), the authorities expand the money stock to generate a higher aggregate demand schedule (AD1) that intersects the new long-run supply schedule at a point consistent with the old price level. This expansion of spending raises the demand for labour to LDI, and so causes the economy to ‘ride up’ its new, short-run labour and aggregate supply schedules to equilibrium points (e) involving higher-than-natural levels of employment and output. As in the case of a pure spending shock, things return to normal once the short-run aggregate supply schedule adjusts. Thus, attempts to stabilise the price level in the face of productivity shocks themselves become a source of disequilibrating monetary misperception effects that would be avoided if the price level were simply allowed to adjust along with changing unit production costs.

    He refers to the graphs in the book. The book is free. You could donwnload it and check out the argument.

    Stabilising spending beats stabilising price level.

  20. Gravatar of kebko kebko
    4. September 2013 at 23:23

    I’m torn, because it seems perfectly reasonable to me, remembering back to 2007, that the deflationary liquidity crisis began well before late 2008, as Steve Roth outlines above.

    But, on the other hand, with wide swings in interest rates, home prices can change significantly – either compared to rent or compared to monthly mortgage payments. So, it seems like changing home values should be treated more like changing stock market values, which shouldn’t necessarily lead to a direct reaction in monetary policy.

    But, on the other hand, if it was common practice for households to leverage up their stock market holdings 5 to 1, then maybe we would need to be more concerned about the stock market’s effect on spending. So, it seems like housing would need to be treated differently, but I’m not sure that the inflation measure is the most appropriate avenue for it.

  21. Gravatar of lxdr1f7 lxdr1f7
    4. September 2013 at 23:33

    I looked at the book you referred to. I don’t see the difference. You end up at the same level of output if prices or ngdp are targeted just at a different price level.

    I dont see what the “disequilibrating monetary misperception” actually is.

  22. Gravatar of Links for 09-05-2013 | The Penn Ave Post Links for 09-05-2013 | The Penn Ave Post
    5. September 2013 at 00:00

    […] Thoma Reply to David Andolfatto – TheMoneyIllusion A reply to Sumner – MacroMania Second reply to David Andolfatto – TheMoneyIllusion The magic fairy NGDP wand – […]

  23. Gravatar of J.V. Dubois J.V. Dubois
    5. September 2013 at 00:44

    I do not know what to think about this debate. The first and most important piece of information is that if last recession showed us anything it is that inflation (or absence of thereof) is not necessary condition deep demand side slump.

    Look at this graph:

    Is there any sane economist that would say that if and only Greece had Central Bank targeting the price level path of 3.3% increase instead of 3.1% it would have full employment instead of staggering 30%? Or maybe, just maybe the recession in Greece has something with the fact that nominal incomes are now 20% below what it was in 2008 and 40% below trend?

    Price level path and nominal income path ARE SIMPLY NOT the same things.

  24. Gravatar of J.V. Dubois J.V. Dubois
    5. September 2013 at 00:55

    llxdr1f7: “”A stable inflation rate is bad for the welfare of the economy …. This seems completely counterintuitive. Is there any evidence for this? What is the reasoning behind this?”

    Look here:

    Short summary: inflation targeting or PLT are bad because they prevent optimal risk sharing of real shocks to real GDP with nominal contracts. This is a staple advantage of NGDP level targeting – that it is not sensitive to supply side inflation.

  25. Gravatar of Vaidas Urba Vaidas Urba
    5. September 2013 at 01:51

    Karl Smith joins the debate:

  26. Gravatar of lxdr1f7 lxdr1f7
    5. September 2013 at 02:32

    J.V. Dubois

    “models of representative consumers miss (i) the well-known distributional effect that borrowers and lenders are affected differently when the price level differs from their expectations,”

    What if the price level target is stable (0%) inflation and all people directly interact with the monetary authority?

  27. Gravatar of Geoff Geoff
    5. September 2013 at 02:45

    Isn’t it interesting that Sumner, Andolfatto, and their followers, are grounding their entire advocacy for a coercively imposed, monolithic central bank with such broad sweeping powers, that affects the lives of hundreds of millions of people, on the basis of selecting “trend lines” on a chart, that they all readily admit are arbitrary; that one is in the last resort just as good as any other?

    And it doesn’t even give them a moment of pause either. I do believe they are bordering on insanity.

  28. Gravatar of lxdr1f7 lxdr1f7
    5. September 2013 at 03:00

    J.V. Dubois

    “Under successful IT or PLT, borrowers absorb the RGDP risk so that the lenders don’t have to absorb any RGDP risk”

    But that is offset by the opportunity cost that lenders risk. If RGDP goes up lenders could of made higher returns.

  29. Gravatar of J.V. Dubois J.V. Dubois
    5. September 2013 at 03:30

    lxdr1f7: “But that is offset by the opportunity cost that lenders risk. If RGDP goes up lenders could of made higher returns”

    We are hijacking this thread with unrelated discussion. Please go and read the article again. The point is that under PLT or IT the real value of the loan at the time of payment is guaranteed. So if both parties (borrowers and lenders) are equally risk averse (that is assumption) then monetary policy that forces no risk on lenders and all risk on borrowers is Pareto inefficient.

    Please note that it does not matter even if this excess risk that is shouldered by borrowers is incorporated in the lower (nominal) interest rate so that even lenders have to pay their share for the excess risk taken by borrowers. It is still Pareto inefficient from welfare analysis point of view.

  30. Gravatar of lxdr1f7 lxdr1f7
    5. September 2013 at 03:59

    Ngdp targeting is pareto inneficient in terms of lost income or opportunity cost if rgdp varies from expected.

  31. Gravatar of J.V. Dubois J.V. Dubois
    5. September 2013 at 05:15

    lxdr1f7: If RGDP varies unexpectedly this poses a risk. Since utility function is quasi concave the Risk is costly. Therefore it mattes how the risk is distributed among persons involved in contract. If their risk aversion is identical then the best way to distribute risk is 50/50 – imagine that this would in real world scenario mean that both parties (borrower and lender) will end up estimating $500 cost to the risk with total cost of $1000. However if the whole risk is forced upon one party he may say that the risk is too much and that he values it as $1200. Therefore this $200 of excess cost to risk compared to optimal distribution of risk is Welfare loss.

    PS: I would like to say that I personally do not consider the effect of aforementioned Pareto inefficiency as what we should care about. We may talk about it if price level targeting and NGDP level targeting have EXACTLY the same impact on macroeconomic stabilization (Andolfato’s assumption). I strongly disagree I am convinced that NGDP level targeting is far superior to stabilize macroeconomy, which is far more important from welfare point of view than ineficiency from risk aversion.

    However even if one assumes that efect of both PLT and NGDPLT are identical on macroeconomic stability there is still an argument for going with NGDPL targeting.

  32. Gravatar of Brian Donohue Brian Donohue
    5. September 2013 at 05:24

    Good comment, Morgan.

  33. Gravatar of lxdr1f7 lxdr1f7
    5. September 2013 at 05:35

    “I strongly disagree I am convinced that NGDP level targeting is far superior to stabilize macroeconomy, which is far more important from welfare point of view than ineficiency from risk aversion.”

    I see price targeting and gdp targeting as the same in the end. Both real cost and opportunity cost have an effect on both parties.

  34. Gravatar of TallDave TallDave
    5. September 2013 at 11:21

    That’s not to say Obama has not moved us 10% or 20% in the direction of France, and raised our natural rate of unemployment by 0.5% or 1.0%. I think he has. But surely it’s now pretty clear that much of the problem in 2009 was inadequate AD.

    That sounds about right. I’ve made similar arguments to people on the right.

  35. Gravatar of ssumner ssumner
    6. September 2013 at 16:44

    Noah, Good question. My sense is that productivity shocks don’t create high frequency movements in the unemployment rate. Or else if they do, then they interact with monetary policy errors. There is a clear gap in jumps in the unemployment rate. You don’t see jumps in the 0.7% to 2% range, which is really weird. By the way, in other countries you do.

    So then what’s going on in 2009 if prices are on target?

    Let’s assume that a gradual price shock (higher oil prices) could be absorbed by the labor market with sound monetary policy. But instead you get bad monetary policy. In 2009 you do get a sharp fall in inflation, indeed you get outright deflation. So in that year my claim works. Over the longer term prices should have been rising at above 2% given the bad productivity numbers and the high oil prices since 2007. So by rising at 2% on average (actually less with accurate price indices) it allowed a sharp fall NGDP, while still keeping prices close to the target path in the long run.

    That’s probably confusing, so here’s another way of making the same point. Given productivity and oil prices, we should have had say 2.5% inflation since 2007. Instead we’ve had perhaps 1.6%. Given a fairly flat SRAS, that’s enough to allow a sharp fall in both NGDP and RGDP.

    Small productivity shocks usually don’t cause recessions because the Fed usually accommodates them. But I need to give the issue more thought, that’s just my first reaction.

    Thanks Steve.

    Morgan, The proximate cause was tight money, but the housing crisis obviously played a role in us getting to that position.

    Mark, Excellent Svensson quote.

    lxdr, Check out my defense of NGDP in National Affairs.

    JV, Good point.

  36. Gravatar of lxdr1f7 lxdr1f7
    6. September 2013 at 23:42


    Assuming NGDP targeting is superior to PLT or IT you must also consider the effectiveness of the mechanism utilized to implement monetary policy. Do you agree that the central bank only directly interacting with banks as opposed to the overall economy is distortionary?

  37. Gravatar of ssumner ssumner
    7. September 2013 at 05:33

    lxdr, I favor the Fed directly interacting with bond dealers, not banks. It actually doesn’t matter very much how money is injected, as long as it isn’t highly wasteful, such as buying real goods or highly risky assets.

  38. Gravatar of jj jj
    7. September 2013 at 18:20

    Say you use PLT or NGDPLT with very long target forecast periods — such as the price level, or NGDP level, 20 years hence. Then make the reasonable modification that the NGDP growth rate need not be constant, but can incorporate a 20-year moving average of productivity (or maybe forecast productivity?).

    Aren’t PLT and NGDPLT then functionally equivalent?

  39. Gravatar of lxdr1f7 lxdr1f7
    7. September 2013 at 19:19


    The price of money for primary dealers is not the price the broader economy is willing to pay. Also the broader economy is not able to access money because the banks are blocking.

    If the broader economy directly interacted with the fed the price for fed funds (rates) would be higher and the demand for funds relative to other assets, comsumption and deleveraging would be alot lower. Therefore the economy would be growing.

    All sectors in the economy should have a proportionate influence. Therefore an effective system of monetary policy would just supply fed funds to the broader economy according to an ngdp target or IT.

  40. Gravatar of ssumner ssumner
    8. September 2013 at 05:44

    JJ, I don’t think so, as the productivity adjustments are done with a long lag.

    lxdr, People can get money from ATMs whenever they please, banks are not blocking that process.

  41. Gravatar of lxdr1f7 lxdr1f7
    8. September 2013 at 07:09

    Banks limit the amount of deposits in circulation which account for most of the money supply.

    People think they have US dollars in their bank as opposed to deposits which are claim on US dollars. This is why people use commercial bank deposits instead of USD’s.

    If people could directly deposit reserves directly with the fed they would because the fed can provide cheaper, safer deposit services than commercial banks.

  42. Gravatar of 123 123
    9. September 2013 at 01:07

    John Taylor joins the debate:

Leave a Reply