The 2009 NGDP shock keeps looking worser and worser
I’m reluctant to even do this post, as the GDP figures keep getting revised downward, suggesting my former comments are “inoperative.” Fortunately for me the revisions keep strengthening my argument. First the BEA said NGDP fell about 2% between 2008:2 and 2009:2. Then last year the figures were revised sharply lower, especially for 2008:3, which showed that the recession got much worse before Lehman. NGDP fell 3% between 2008:2 and 2009:2. Now we have another revision, and the fall is nearly 4%! (minus 3.85% to be precise) What was already the worst AD collapse since 1938 just got even deeper.
What’s up with the BEA? It isn’t just that they are revising RGDP data from years ago (I suppose a new price index formula could explain that), but they are also sharply revising nominal GDP data from years ago. They are still getting new reports of nominal output! Even more bizarre, the nominal numbers are being revised downward by massive amounts, more than $100 billion. So they are getting new reports that nominal output for early 2009, which they still thought existed a year later in early 2010, did not in fact exist. I’m pretty sure that I would be horrified by seeing what goes on inside the BEA—I’d rather visit the Oscar Meyer factory in my hometown, to see how hot dogs are made. Employment numbers are probably our best “real” indicator.
So we now have a nice three year data set. NGDP fell nearly 4% from 2008:2 to 2009:2, and has risen at a tad over 4% a year for the next two years. We are up by only 4.1% in 3 years, that’s a little over 1% per year. In other words, per capita NGDP has barely risen in 3 years! To maintain full employment everyone would have had to go nearly three years without a pay rise. But with soaring minimum wage rates, that wasn’t too likely. I’m sure lots of people in government, health care, education, etc, got raises. I did (even with one year of no raise), as did my wife (a scientist.) So with almost no extra NGDP to go around (per capita), we essentially have a game of musical chairs. The lucky ones get pay increases, the other 9.2% are sitting on the floor.
Now I’ve got to revise my papers. I’ve been saying NGDP fell 8% below trend during the contraction, actually it was 9%. And we are 11% below trend over the three year period. Obviously we are never going all the way back, nor would I advocate that. But we are 5% below Bill Woolsey’s quite conservative 3% NGDP target, starting from mid-2008. A target most ultra-conservative inflation hawks would have grabbed in a heartbeat, if offer the chance in mid-2008.
There’s a bit of good news for those pushing the structural problems argument. Previously the real/nominal split during the recovery was estimated at 2.8%/3.9%. Now it’s estimated at 2.5%/4.1%. So we have a tad more stagflation that we thought, but still not very much. More worrisome is the split over the last two quarters, 0.85%/3.4%. That may reflect the effects of the oil shock, but we need to watch that carefully to see how severe the economy’s structural problems really are. I am a moderate on this issue, and believe we have some structural problems in the labor market, but that it’s mostly demand-side. The new figures haven’t led me to revise this view, but if they continue for several more quarters, I will definitely revise this view. Although I won’t revise my monetary policy recommendations, which don’t depend at all on real variables.)
Tags:
29. July 2011 at 08:03
“The new figures haven’t led me to revise this view, but if they continue for several more quarters, I will definitely revise this view. Although I won’t revise my monetary policy recommendations, which don’t depend at all on real variables.)”
This is perfect timing!
Perry will be President. You will agree the problems are structural. And conservatives will finally be willing to target NGDP.
29. July 2011 at 08:36
I’m increasingly upset at the Fed. Forget about skepticism over whether QE works, or fears of too much inflation. We have a small core of Republicans right now who may insist on a US default, and even welcome it, who may have never even gotten elected had the Fed done more in ’08 and ’09.
And even now, as things get increasingly dangerous and bizarre, and after a terrible couple of economic quarters with below target numbers, the Fed sits on its hands.
This is possibly becoming about much more than just preventing a lost decade or lost generation, or worse economically. This is about a real breakdown at multiple levels of government and society. The Fed needs for forget about the rest and just stimulate.
As bad as default might be, a debt deal that has drastic spending cuts will be tragic too, especially without some offset, and may continue to corrode our politics and society while introducing unbearable burdens on millions already suffering more than anyone has since the Great Depression.
The Fed is supposed to have some independence for a reason, and all they do is seem to reflect a very misguided general consensus among establishment types.
29. July 2011 at 08:48
I’m confused. I thought you wanted us to return to NGDP trend, and that was your monetary policy recommendation… Also, didn’t the report show numbers that strengthened the case for more AD? I read that main causes in lackluster growth were declining government spending and consumer spending (business having increased their spending).
I suppose I would just like to better understand what the case for more structural problems than we thought we had before…
29. July 2011 at 09:16
Scott,
Just thinking a bit of some Sumnerian analysis….since we can not really observe directly how tight monetary conditions are you have argued that we should be using market data. In that regard I have noticed that recently TIPS breakeven inflation rate have been inching up and it significantly higher than prior to the talk of QE2 started (and later was implemented) and the dollar has been weakening recently.
Hence, despite the weak US macroeconomic data the market does seem to be thinking we are heading for deflation. So yes, the data is very weak and NGDP is well below where it should be if we compare to a 3 or 5% trend going back to prior to the crisis, but can we say that monetary conditions are becoming tighter? No, rather we would have to say that rising (TIPS) inflation expectations and a weaker dollar is indicating easier monetary conditions. Or can we?? To me the answer is that what we are seeing is that the market perception of the US default risk is on the rise more than anything else. Nick Rowe recent told us that rising default risk should be analysised as increased inflation expectations – and in an ISLM-setup with a liquidity trap actually is expansionary. So maybe you are too negative?? I don’t think and worry as well, but…
Furthermore, if stocks decline and TIPS inflation expectations are increasing isn’t that an indications of worries in the market about RGDP rather NGDP??
And finally is it the case that when weak US data leads to an increase in inflation expectations and a weaker dollar then it is in fact an indication that the Fed is regaining some credibility in terms of responding to deflationary risks? If the market exepcted the Fed to be passive to the risk of deflation then weak data should have lowered (market) inflation expectation and the dollar should have strengthened.
You know I agree with you 95% of the way, but what is best – market reaction to data or the actual data?
29. July 2011 at 09:18
“Hence, despite the weak US macroeconomic data the market does seem to be thinking we are heading for deflation.”
Should of course be…
“Hence, despite the weak US macroeconomic data the market does NOT seem to be thinking we are heading for deflation.”
29. July 2011 at 09:18
Morgan, No new Texans!
Mike, The problem is that they are the establishment.
Marcelo, Yes, we need much more AD. What I meant was that the RGDP/NGDP is estimated to be less RGDP, and more inflation. That strengthens the argument that there are structural problems, although needless to say I think the main problem is demand.
29. July 2011 at 09:21
Lars, I’m 100% for market data, but don’t trust inflation data. I favor targeting market NGDP growth expectations. So while I agree that the oil shock has recently raised inflation a bit, I think it has lowered market expectations of NGDP growth, which is reflected in falling long term nominal bond yields. (I focus on the five year yield, down to 1.41%. That suggests very slow NGDP growth ahead.)
29. July 2011 at 09:23
Lars, I should add that I see the lower dollar as reflecting a weaker economy, not easier money.
29. July 2011 at 09:31
Perry is your best shot a being handed a cushy Fed job.
You are going to love him.
29. July 2011 at 09:31
Ok. Thank you for the answer. I think my main concern is that I thought you wanted the NGDP to go back to trend, and you say that this is a bad idea. I felt like the best solution would be to return NGDP back to trend and then enforce NGDP trend targeting, and let whatever short run inflation do what it will. Perhaps I misunderstood your previous stance.
If the Fed were to go back to return us to our previous trend NGDP what do you think is the worst that would happen? If expectations are set to trend NGDP and EMH holds, then I don’t see how inflation gets out of control…
Again, thanks for being so responsive to comments, as an undergrad econ student I find this blog helps apply a lot of theory to the real world, and makes more sense to me than most of everything else out there! Keep up the good work =)
29. July 2011 at 09:36
Marcelo, if all undergrad econ students read this blog I am pretty sure that we could be optimistic about the future…and damn I wish some central bankers could read and learn a bit as well…
29. July 2011 at 09:45
The growing RGDP/GDP spread is worrying. The U.S. political system probably isn’t capable of making the reforms needed to fight a stagflation problem. A demand shortage can always be fixed by better monetary policy, a supply shortage means we are doomed.
29. July 2011 at 10:44
JPIrving, I don’t think the US has a real serious supply side problem. However, as somebody who in my day job works with Emerging Markets analysis I myst say that US politics increasingly looks Emerging Markets’ish…or real the kind of political crisis the US going through these days is WORSE than what you are seeing in most EM countries – whether in Central and Eastern Europe, LATAM or Emerging Asia…not too impressive. To me the markets are reacting more to the political crisis rather than to the actual economic data – and I guess the ratings agencies will do so as well soon…
29. July 2011 at 12:37
“Obviously we are never going all the way back, nor would I advocate that.”
So what’s the optimal response right now? What should the Fed do now to make up for the amount we’ve fallen below trend?
Is there a fully spelled out model you’d look to for this answer?
29. July 2011 at 14:42
Morgan, I once turned down a cushy Fed job at the NY Fed. It would gave given me a 75% pay increase.
Marcelo, I probably didn’t fully answer your comment. In general I do favor going back to trend. But there comes a time when there is so much “water under the bridge” that it’s no longer realistic. Japan would have to have 15% inflation to make up for all the deflation. But by now wages have mostly adapted to the mild deflation, and probably just somewhat higher NGDP growth would do the job. My point is level targeting should be the official policy. If they had told markets in 2008 that any 2009 shortfall would be made up in 2010, the recession would have been much milder. But we’re so far behind now that I don’t know if it makes sense to shoot for full trend reversion. Wages have partially adjusted. It’s a judgment call.
JP Irving, We’ve got problems, but it’s partly the oil shock in the last two quarters.
Charlie. It’s up to the Fed to set a goal. I can’t tell you exactly what they should do, unless I know their goal. I’d probably just shoot for a few years of 7% NGDP growth, and then go to their long run trend goal. It doesn’t have to be 5%, but that’s been their implicit target in the past.
29. July 2011 at 16:09
Scott,
I guess what I’m asking is, if I were to take over the Fed tomorrow and think the Fed should have been targeting 5% NGDP, how should I exit. Isn’t there some parameterized model that tells me to go 7% for x years and then back to 5% or 10%, 9%, 7%, 6%, 5%?
29. July 2011 at 18:07
Sorry Charlie, There is no model to tell us what to do, because the Fed never announced their policy. If they had an announced policy, we’d have a model to work with.
29. July 2011 at 19:07
“Sorry Charlie, There is no model to tell us what to do, because the Fed never announced their policy. If they had an announced policy, we’d have a model to work with.”
In what way does it matter? We have a pretty good where expectations were grounded even though the Fed never said they were targeting 5% NGDP or inflation targeting 2%.
What are we missing in the model?
29. July 2011 at 22:06
Scott,
You may well be looking at a situation where NDGP targeting would not result in repairing RGDP, but at an economy that has “permanently” shrunk its productive base and hence will just see certain prices go up (at least for quite a while). Those unemployed people need capital (and maybe new skills) in order to generate the output/capita they used to produce. It seems to me that that is not (yet) happening. They do not do that by just being on the dole (with or with some informal job on the side), or staying in whatever country they went back to.
So I guess that next to NDGP targeting, someone has to find out how to deal with these people -and the ones soon returning from unaffordable military work- Or not?
30. July 2011 at 01:21
WSJ: “Fed’s Bullard: Rising Inflation Makes New Stimulus Unlikely”
And in related news:
“The Federal Reserve Bank of Cleveland reports that its latest estimate of 10-year expected inflation is 1.83 percent.”
Bullard’s President of the St Louis Fed. Do St. Louis and Cleveland talk to each other? They’re not that far apart. Maybe they should get together for lunch.
30. July 2011 at 02:00
[…] Source […]
30. July 2011 at 04:02
The BEA makes a truly noble effort…even if you don’t want the details of how the sausage is made. In the case of consumer spending, they incorporated the most recent results from annual (mandatory) establishment surveys on services and retail trade for 2009. They won’t have the ‘final’ annual results for 2010 sales until next year’s annual revision…You are correct the best, real-time read we have on the economy is the unemployment rate.
30. July 2011 at 10:26
Charlie, Even if we accept that they implicitly target NGDP growth at 5% (and right now they aren’t) then we need to ask whether they are doing level targeting or rate of growth targeting. In other words, to give you advice on where you should go next year, I’d need to know where you plan to go in 5 years, or 10.
Analogy. Say you are driving to California, and want to know whether to go through Denver or Phoenix. I’d ask. “Are you going to LA or the bay area?”
If they are doing level targeting, I’d need to know what trend line they are aiming at. If you want me to recommend a trend line, I’d probably say something like 7% for two years, and 5% thereafter. But there’s no magic formal as to what’s best. On the other hand if I knew the trend line they are shooting for, I could give much more accurate advice.
Rien, You might be right, but it would be the first time in American history this happened. More likely you are 20% or 30% right, but we’d still get lots of RGDP growth. Lots of businesses are complaining about lack of demand. They say they’d hire more workers if the sales were there. That includes basic services like restaurants, retailing, etc.
Jim Glass. Good point. The 10 year TIPS spread is higher than that, but it has technical flaws (like a lag in indexing) that the Cleveland Fed tries to correct.
Claudia, Thanks for that info. Care to hazard a guess as to why in the UK the NGDP data comes out three months after the RGDP data?
30. July 2011 at 13:11
“If they are doing level targeting, I’d need to know what trend line they are aiming at. If you want me to recommend a trend line, I’d probably say something like 7% for two years, and 5% thereafter.”
I understand that, but now I’m confused. You’ve been recommending level targeting the entire time, right? So why not advocate now a path that gets us back onto the trend the Fed abandoned in 2007?
I guess if I could clarify where I think you are answering a different question than I’m asking. I’m not asking how Scott Sumner would target NGDP to achieve Ben Bernanke’s objectives. I’m asking how Scott Sumner would target NGDP to achieve Scott Sumner’s objectives, if he were to control U.S. monetary policy. And if the answer is, 7%, 7% then 5% long term, why is it optimal to not try and get back on the pre-recession NGDP trend?
Again, just trying to clarify and understand.
30. July 2011 at 19:30
Scott,
Thanks and your following posts is quite useful in this respect.
Just this, you said:
Lots of businesses are complaining about lack of demand. They say they’d hire more workers if the sales were there. That includes basic services like restaurants, retailing, etc.
When I wrote my comment, I was tempted to mention that type of jobs too, for the reason that they would have a much lower than average productivity and as such would be a drag on long term growth. Better for current GDP than the dole, but not indicative resumption of growth.
Anyway, we all agree that the current situation is dangerous. Maybe the controversial European bailouts will be just the thing (combined with Japanese reconstruction which had a much slower start than Kobe in the 1990s, but is a bit on hold pending resolution of the fossil or nuclear question) that keeps the world from turning a cold into pneumonia..
30. July 2011 at 19:31
The BEA published its first estimate of GDP for 2011:Q2 on Friday…one month into Q3. However that is a preliminary estimate that is subject to a number of revisions as they fold in better and better source data. For example, the first three estimates of most goods spending in the NIPAs are based on voluntary survey responses from a sample of retail trade establishments. Two years later the BEA will be able to incorporate the results of a mandatory census of these establishments. Clearly a comprehensive census is more definitive than a sample survey, but it takes time to collect the census. Unfortunately, revisions appear to be cyclical, so the measurement error in the real time data is probably not mean zero.
31. July 2011 at 06:27
Charlie. Level targeting only works if you do it in a timely manner. You need to immediately try to recover back to the old trend line. Otherwise why not go back to the 2000 trend line, or 1990? In the model there is nothing special about 2007.
Rien, Those service jobs are better than unemployment until the “good jobs” come back. I didn’t know that about Japan. Actually they did a poor job after Kobe, so if this is worse then they are in big trouble.
Claudia, Thanks for the info. The bottom line is that RGDP data is not very reliable. The best you can say (and what’s necessary for NGDP targeting) is that the errors are random.
I’d look to jobs data, not RGDP data, as being the better real time indicator. There’s no doubt in my mind that RGDP grew more than 0.4% in Q1. The unemployment rate fell from 9.8% in November 2010 to 8.8% in March 2011. That doesn’t happen with 0.4% growth.
31. July 2011 at 07:15
I would agree in real time that the signal from the labor market is probably more reliable than from the NIPAs, but all measures of economy are noisy…especially during a massive cycle… so you might want to hold onto your ‘doubts’ for awhile.
1. August 2011 at 08:50
Claudia, I agree.