Farewell to Ben Bernanke
Morgan sent me this WSJ piece by Austan Goolsbee:
Bravo for Bernanke and the QE Era
Admit it: The Fed’s loose money policy worked. Now the challenge is how to manage the tapering.
. . .
Think back to the days before the 2008 crisis or recession. If confronted with the scenario that would follow””five years of GDP growth of only around 2% a year, five years of unemployment rates around or above 7%, core inflation consistently below 2%””the near-universal response of economists would have been for the Fed to cut interest rates.
So Goolsbee views this as success, but never tells us why. He never explains what the Fed should be targeting, or by what criterion he judged the policy a success.
I believe Goolsbee is representative of the economics profession circa 2014.
His speech’s audience at the nation’s largest gathering of economists looked even more comical than normal with their snow boots and hat-head, but they gave his remarks a standing ovation. It was quite different from the icy reception he has grown accustomed to from his critics (including several on this page) over the past 3½ years during what could be called the QE Era.
The Fed did poorly over the past 5 1/2 years. But perhaps Bernanke does deserve a standing ovation. Everything’s relative. As far as I can tell Bernanke outperformed the profession, which is all we can really ask of a policy leader. Here is some evidence:
1. Polls show the profession was generally either equally hawkish or more hawkish than Bernanke. Before QE3 and the forward guidance decisions of late 2012 hardly any economists thought policy was too tight. Many thought it too easy.
2. However Ben Bernanke thought it was too tight. In mid-2012 he didn’t even have the rest of the Fed behind him even though it was mostly Obama appointees. During the summer of 2012 he had to twist arms and schmooze Fed officials to persuade them to sign on to a policy that successfully prevented the austerity of 2013 from driving the US into a double-dip recession.
3. He outperformed the most comparable other large central banks, in the eurozone and Japan. And not by a small amount, he massively outperformed.
4. Based on their recent public statements, either Greenspan or Volcker would have been complete disasters. Trichet clones. And remember that Greenspan and Volcker are two of the most successful Fed chairmen in history. Thank God we had Bernanke and not either of those two. Don’t know if age was a factor, but they had clearly lost it.
[Last time I made that comment I got accused of ageism. What nonsense! How do you think Lebron James will be doing at age 85? Bernanke served from age 52 to 60, peak years for a monetary economist. BTW, I started blogging at age 53, and am currently 58. If I live to be 85 I’ll be rambling incoherently about income, inflation, and interest rates all being completely meaningless. Oh wait, I do that already.]
5. Bernanke had studied both the Great Depression and the Japanese “liquidity trap” and hence was better prepared than almost any other economist to deal with the crisis of 2008. From a market monetarist perspective he did poorly. But market monetarism is a tiny fringe group that’s completely out of the mainstream. There was absolutely no way that Bernanke could have implemented the MM agenda even if he had wanted to. Judging the Fed chairman by this criterion would be like judging a President of the US on the assumption that he was a dictator freely able to enact the preferred agenda of a Bryan Caplan or Matt Yglesias. You judge people according to their marginal product. Bernanke did better than most would have done in his shoes. And for that he deserves thanks, and an enjoyable retirement.
If the economics profession had been solidly market monetarist in 2008 then I’m confident that Ben Bernanke would have gladly implemented NGDPLT. The economics profession never gave him the support he needed to be more aggressive. The profession failed us, the Fed was just a symptom.
OK, start hammering me in the comment section for being soft on Bernanke. I don’t care.
Tags:
30. January 2014 at 19:55
If only we had listened to him and Paul Krugman:
“…federal fiscal policy has turned quite restrictive; according to the Congressional Budget Office, tax increases and spending cuts likely lowered output growth in 2013 by as much as 1-1/2 percentage points. In addition, throughout much of the recovery, state and local government budgets have been highly contractionary, reflecting their adjustment to sharply declining tax revenues. To illustrate the extent of fiscal tightness, at the current point in the recovery from the 2001 recession, employment at all levels of government had increased by nearly 600,000 workers; in contrast, in the current recovery, government employment has declined by more than 700,000 jobs, a net difference of more than 1.3 million jobs.”
Read more: http://www.businessinsider.com/bernanke-on-fiscal-policy-2014-1#ixzz2rwde3PwA
30. January 2014 at 19:58
The WSJ editorial board also wrote a farewell to Bernanke (warning: more hostile).
http://online.wsj.com/news/articles/SB10001424052702304691904579346484179872324
30. January 2014 at 20:09
As I’ve said before, the fun guy to listen to from 2008, maybe not not on MP, but on what the Fed cared about… was Greenspan. He became the Fed’s id.
And he spent those years just harping on HOME PRICES and keeping PRIME BORROWERS prime.
Hopefully, the Fed’s new id will be Ben.
30. January 2014 at 20:20
Could the low reported PCE inflation rate (lowest since 2009) explain why the stock market rose 1.13% today?
http://soberlook.com/2014/01/pce-inflation-rate-lowest-since-2009.html
30. January 2014 at 20:30
Best farewell ever.
30. January 2014 at 20:40
Speaking of age, Janet Yellen is currently 67 years old. At such an advanced age, could she have trouble pushing FOMC members to support her positions?
30. January 2014 at 20:54
Travis, called me sexist, but I find that women who’ve been that successful for that long are usually pretty good at getting their way. Yellen would have faced a lot more obstacles in her time than Bernanke. I think she will be pretty good at banging heads together.
30. January 2014 at 22:15
Yeah, you’re too soft on Bernanke. He didn’t need to convince the economics profession. He just had to convince 6 or so members of the FOMC. If he wanted more FOMC members who were more favorably inclined to his (your?) way of thinking then he could’ve suggested people for Obama to appoint and made clear the urgency of the matter. Those vacancies probably would not have gone unfilled for as long. Kocherlakota was persuadable, obviously, in retrospect. Maybe he would have responded earlier with the right arguments. Bernanke could really have revived and embraced his prior recommendations for Japan and explained why they applied to our predicament instead of partially repudiating them.
Overall he was incredibly timid. It was his choice to attempt to run the Fed by consensus. A stronger personality might have tried to create a working majority, confident enough his policies would succeed that they would eventually be accepted by the dissenters.
I agree that he was probably better than most of the people that might have been in the position at the time. Better than Glenn Hubbard and Larry Summers certainly, although Mankiw might have been considered and he favored a 3% inflation target in 2009. Of course, seeing as what happens to everyone when they are in power he probably would have been cowed by the Borg into towing the line. Still, even with all of the limitations Bernanke could’ve done a lot better. What about simply having implemented the current style of QE based on inflation and employment data way back in 2010?
30. January 2014 at 22:58
“he didn’t even have the rest of the Fed behind him even though it was mostly Obama appointees”
With economic advisors like Goolsbee it should not be surprising that Obama failed to appoint people on the BOG who would take the Fed’s mandate to seek maximum employment seriously.
30. January 2014 at 23:28
I would have stood up for Bernanke as he struck as a sincere and dedicated public servant and stayed in a tough job for a long time. Perhaps his nature is timid, also.
We can disagree without bring disagreeable.
That said, he should have printed a lot more money.
31. January 2014 at 00:24
“There was absolutely no way that Bernanke could have implemented the MM agenda even if he had wanted to.”
One did no have to be a market monetarist to push for a monetary policy that would more closely comply with the Fed’s mandate to seek maximum employment. I would conjecture that if Chritina Romer had been Chair, she would have tried to get the FOMC to implement a more expansionary monetary policy.
Bernanke until QE3 was implemented did appear to be excessively timid. At the time I called him the Hamlet of monetary policy.
31. January 2014 at 01:18
He’s certainly been a magnificent support for those who invested in equities and debt, and to reckless mismanagers of top financial firms. Surely the economy would have crashed harder in 2009 without him, but would it have bounced back faster?
Well, since his and the Bush administration’s bailouts were of such a scale that would be politically almost impossible to repeat – indeed, to some extent such bailouts are prohibited by law – we will likely find out next time the banking system crashes. JPM doubling Dimon’s pay sure looks like a step in that direction.
31. January 2014 at 03:30
Scott, is it not your position that the Great Recession was caused by the Fed? Under the Chairmanship of Ben Bernanke? He gets off scot-free for this? What is a chairman anyway?
31. January 2014 at 03:38
The US massively out-produced other countries in oil production growth.
On a related note, Shell yesterday announced a 20% cut in capex (at $107 Brent); Hess will have cut capex by 30% in the 2012-2104 period. These cuts will spread to the rest of the publicly-listed operators, in multiple rounds, totaling about $80 bn during the next two years. This is likely to lead to a loss of 200,000 jobs, primarily in Houston, Aberdeen and Stavanger (Norway).
31. January 2014 at 03:42
Two notes on Trichet:
– Trichet’s response to Lehman was much better than Bernanke’s. Trichet’s policy innovations were extended by Draghi. As a result, Greece’s default did not become Lehman II.
– Trichet’s interest rate policy in 2011 was worse than Bernanke’s. The outcome of Trichet’s 2011 policy is a strong argument against Taylor rules.
31. January 2014 at 03:44
On a further related note, due to an unfavorable court ruling, Shell has decided not to drill in Alaska this year. The odds of Shell abandoning Alaska have risen to perhaps 50% or more. This, in turn, implies the exit of Alaska from the oil business in the 2020-2030 time frame and the material destruction of the state’s economy.
See more in my Oil & Gas Journal article: http://www.ogj.com/articles/print/volume-111/issue-11/exploration-development/alaska-s-oil-crossroads-lucrative-ocs-prize-and-taps.html
To quote from the article: “The sobering truth is that Alaska is one bad board meeting at Shell away from seeing the end of its oil age.” That’s what’s on the table.
31. January 2014 at 05:27
Scott,
If fracking technology had not worked do you believe it would have been possible for the Fed to have done the QE it did do? What would be the political ramification of $5 or $6 a gallon gasoline? What would be the economic impact of such?
BTW, US oil production is up 40% from 2008. Without this unexpected supply increase the QE program would have been at best short lived. Which then begs the question of whether the economic growth we have had is due to MM or a result of supply side innovation. Most likely the credit should go to the engineers and not the economists.
31. January 2014 at 05:28
Tom, You forget that Krugman suggested 2013 was a test of that theory, and the Keynesians lost.
Travis, Don’t know, was there an announcement effect?
Thanks Henry.
Danny, He was a net plus, but not as big a net plus as you or I would have liked.
Ben, I agree.
FEH, Let’s say that 20 of the 7 billion people in the world would have been better chairs than Bernanke. And let’s say Romer is one of them. That’s still consistent with Bernanke being a well above average public servant.
Tom, Don’t see where he’s helped stock investors very much.
Saturos, Not sure he gets off “scot-free” (was that a pun?) I did say the Fed performed poorly over the past 5 1/2 years. I stand by all the specific assertions in the post.
Vaidas, I’ll defer to your expertise on Trichet in 2008.
31. January 2014 at 05:31
Dan, You do realize that the US economy grew in earlier decades, don’t you? And that was without any help from fracking. Growth is the norm. All we’ve done is grow at 2% to 2.5% in recent years, which doesn’t require any special “fracking” explanation.
31. January 2014 at 05:32
Given the seriously excessive unemployment that stemmed from the crisis, and given that the source of fiscal stimulus i.e. Congress has consisted for several years of a bunch of economic illiterates engaged in a punch up, it really didn’t take a genius to work out that monetary policy need to come to the rescue.
If the most diehard Keynsian had been Chairperson of the Fed, they’d have done the same as Bernanke (hopefully).
31. January 2014 at 05:47
“OK, start hammering me in the comment section for being soft on Bernanke. I don’t care.”
He caused the GFC, Scott. You’ve said so in so many words too. The fed mistakenly misunderstood a change in relative prices in the commodity markets during the middle of the decade, took that as inflation, tightened far too much and then was too slow in realizing the damage caused. Ben was an absolute disaster at the Fed.
As for comparing Volcker or Greenspan.. perhaps those two would not have tightened so much and we avoided the GFC.
31. January 2014 at 06:08
Scott. Nicely balanced comment on Bernanke. However … the Fed still caused the Great Recession and can still cause another.
Looking forward this last FOMC has tightened monetary policy,check out the equity and bond markets around the time of the announcement. (You know, like you used to, before going soft 😉
At the December meeting they loosened by offsetting the start of the taper with some nice long term guidance. That nice long-term guidance is now rather fuzzy, and the lack of concern about current macro-conditions was a surprise as was the unanimity of the decision to march onwards with the tapering. The consquences are continuing to reverberate.
It almost seems as if the Fed really didn’t like to doing the 2013 monetary offseet and can’t wait to return to “normal”.
31. January 2014 at 06:19
Scott,
How do you believe QE would have worked politically and economically if domestic US oil production had not increased 40% from 2010 – 2013?
Do energy supply shocks matter or not? If you say no then how would you rewrite the economic history of the 1970s?
31. January 2014 at 06:44
Great post. Before I discovered this blog, I considered Bernanke among the greatest central bank chairmen of all time. Standing up to those hawks and preventing another Great Depression. Then I started reading your posts and for the first time found intense criticism of Bernanke from the other side.
While I considered the criticism valid, I always thought Bernanke was doing a better job than almost anyone else would have done. I don’t know if someone else would have started loosening earlier, say in 2007. It’s possible, and I actually think Greenspan would have been looser than his comments indicate.
So I say Ben Bernanke did a good job, and should enjoy his retirement. But he should come out for NGDPLT soon.
31. January 2014 at 07:51
Scott,
Have you seen this yet?
Ramesh Ponnuru:
“…There’s another view of the Fed’s role in the crisis, though, that has been voiced by economists such as Scott Sumner of Bentley University, David Beckworth of Western Kentucky University and Robert Hetzel of the Richmond Fed. They dissent from the prevailing view that the Fed has been extremely loose since the crisis hit. Instead, they argue that the Fed has actually been extremely tight, and that when its performance during the crisis is measured against the proper yardstick, the central bank emerges as the chief villain of the story…”
http://www.bloomberg.com/news/2014-01-30/so-long-to-bernanke-and-his-tight-money-fed.html
31. January 2014 at 07:58
Dan W.,
Excellent question!
31. January 2014 at 08:19
Dan W, see this post from Scott on money vs. supply shocks:
http://www.themoneyillusion.com/?p=20114
31. January 2014 at 08:29
Vaidas Urba,
“Trichet’s response to Lehman was much better than Bernanke’s. Trichet’s policy innovations were extended by Draghi. As a result, Greece’s default did not become Lehman II.”
This is hugely debatable, especially since probably the biggest initial response to the crisis by the Fed consisted of nearly $600 in dollar swaps whose main beneficiary was the ECB. So is Trichet supposed to get credit for Bernanke pulling the ECB’s arse out of the fire?
Also it’s hard to forget that the ECB *raised* the MRO rate in July 2008 which happened to be precisely when the financial markets started blinking red alert warning signals. In fact I am of the opinion that the original impulse for the global great financial crisis started in Frankfurt, not Washington, and I’m not alone in that opinion:
http://marketmonetarist.com/2012/07/25/between-the-money-supply-and-velocity-the-euro-zone-vs-the-us/
“Trichet’s interest rate policy in 2011 was worse than Bernanke’s. The outcome of Trichet’s 2011 policy is a strong argument against Taylor rules.”
What possible Taylor Rule are you talking about?
Here’s what the European Commission’s Spring 2011 Quarterly Report on the Euro Area had to say about the output gap and inflation on the eve of the interest rate increases in April and July 2011 (Page 16):
http://ec.europa.eu/economy_finance/publications/qr_euro_area/2011/pdf/qrea1_en.pdf
“..The economic and financial crisis of 2007-2009 has resulted in a large output gap that is only gradually closing. According to the Commission’s autumn 2010 forecasts, the output gap of the euro area reached a trough of -3.8 % in 2009 and is projected to remain sizeably negative for some time, reaching -1.6 % in 2012. For comparison, the OECD Economic Outlook of November 2010 sees the euro-area output gap at -4.9 % in 2009 and -2.7 % in 2012. (13) Yet, euro-area core inflation area has been remarkably stable. From a peak at 2.7 % in March 2008 it has fallen to a trough of 0.8 % in April 2010 and since then gradually climbed back to 1.1 % in February 2011…”
So the ECB chose to raise rates in 2011 despite the fact that the European Commission’s estimates showed the output gap would “remain sizably negative for some time” and despite the fact core inflation was only 1.1% in February 2011.
To be blunt, Trichet is living proof that you should never leave the monetary policy of the second largest currency area on earth to a mining engineer. And although Draghi is a huge improvement, that’s only because Trichet set the bar so low.
31. January 2014 at 08:58
Dan W.
The entire mining and quarrying sector of the US economy has grown from 1.05% of gross value added in 2009Q3 all the way to…1.75% in 2013Q3 (gasp!):
http://research.stlouisfed.org/fred2/graph/?graph_id=141978&category_id=0
If the fracking boom had never occured I don’t think anyone other than the Dakotas and Steve Kopits would have noticed. (Well, I guess everyone needs a hobby.)
Please don’t tell me you too are infected with Petroleum Derangement Syndrome (PDS).
31. January 2014 at 09:09
Dan W. and Mark Sadowski,
This topic deserves more attention because I’m very sympathetic to Dan W.’s argument.
Real growth has increased substantially since 2011. How much of that is due to new oil discoveries and how much is due to monetary expansion?
Inflation was pretty darn high in 2011. Since then, it’s fallen dramatically yet economic growth has accelerated. Isn’t that hard for market monetarists to explain without the new oil discoveries?
31. January 2014 at 09:13
TravisV,
I think that the figures that Mark cites do away with that explanation. The fracking boom is huge news for those directly affected, but it’s not that important.
Also, have we got to the point where 3.2% is “pretty darn high”?
31. January 2014 at 09:14
And why do people keep on asking what the MM view is on something before consulting the NGDP figures? It’s very frustrating.
31. January 2014 at 09:33
Travis V wrote:
“Real growth has increased substantially since 2011. How much of that is due to new oil discoveries and how much is due to monetary expansion?”
The way I look at it, monetary policy will affect the economy’s ability to reach its highest possible rate of real growth.
Under the best possible monetary system, growth would be wholly determined by real factors. To the extent that monetary policy is not optimal, it will detract from whatever real growth is possible. Real factors set the upper bound on what an economy might achieve, but monetary policy will affect its ability to realize that potential.
NDGP growth, on the other hand, is not in any way real. It is completely a function of the supply of money and the demand for money. Those who control the supply of money can have whatever trend level of NGDP growth they want, precisely because it is not real.
31. January 2014 at 09:38
Dan W., Travis V, W Peden,
The issue of how much fracking mattered to economic growth depends on your view of supply elasticity elsewhere in the world, as well as demand elasticity (conservation) in the US.
Fracking has added around 6 million barrels per day to US supply (half natural gas). If you think the ROW would have been able to produce that on demand, that the US would have been able to bid that supply away from China despite the trade deficit, or that the people would have been able to cut energy consumption 20% with no real economic impact, then the Sadowski view is correct.
But if you think supply was constrained, then the Sadowski view is wrong.
I previously pointed out that using mining share of GDP is the wrong approach since fracking has dramatically reduced prices. Arguably mining share would be higher today without fracking. But that won’t stop this zombie from being repeated endlessly, which is why I gave up on the debate.
31. January 2014 at 09:42
Fair review of Bernanke.
I think the really interesting question is… what will the Fed do during the next recession? Have they learned to keep a closer eye on NGDP? Or not.
31. January 2014 at 09:45
TravisV,
This topic has already been savagely pummeled to smithereens.
I’ve had at least three prolonged back and forths with Steve Kopits on this subject here and at Econbrowser (and one with Steve).
Not to be excessively rude, but Kopits reminds me of some conversations I had with a creaky old computer program called Eliza back in the early 1980s. His answers sound responsive but he doesn’t actually mentally process a damned thing you say. And quite frankly, the last conversation I had with Kopits at Econbrowser I thoroughly cleaned the floor with his arse for good measure precisely because I was tired of hearing this nonresponsive BS over and over again.
It’s time to bury the whole “shale oil boom explains everything” zombie myth deep in its tomb and seal it up for good. It’s nonresponsive, it’s incoherent, and quite frankly it’s stupid.
31. January 2014 at 09:49
Michael Byrnes,
I absolutely agree that easier money encourages higher real growth. But if we had easier money, I would expect…..higher inflation.
But we don’t have higher inflation, we have dramatically lower inflation. Hmmmm……
Intuitively, it sure seems to me that monetary policy has shifted the demand curve rightward while “other stuff” has shifted the supply curve rightward substantially in 2012 and 2013.
31. January 2014 at 09:55
If you want to know what would have happened without fracking you need to be clear on the counterfactual.
Maybe Obama would have fast-tracked 100 coal-fired power plants. Then we could grow like China and fracking wouldn’t matter.
Maybe Poland would have invented fracking instead. Then Poland could grow like North Dakota, and the US would be growing more like Europe is.
Maybe there would have been no alternative, and then the economy would be in bad shape.
But if your “no fracking” counterfactual to really just fracking somewhere else (maybe Canada) then the Sadowski answer is reasonably close.
31. January 2014 at 10:10
What would the economy be like if Intel hadn’t happened? The same, we’d all use AMD chips.
What would the economy be like if Moore’s Law hadn’t happened? We’d be a lot poorer today.
The debate about “fracking” needs to be seen in this light. Is the counterfactual an equivalent technology with a different name/location? Or is the counterfactual no substitute technology?
31. January 2014 at 11:13
“The entire mining and quarrying sector of the US economy has grown from 1.05% of gross value added in 2009Q3 all the way to…1.75% in 2013Q3 (gasp!):”
Mark, I’m curious about this and perhaps you have an answer. (Hopefully, an objective one because you seem to have invested a lot in downplaying the fracking effect).
The question is this: In computing the gross value added of the mining and quarrying sector, does this measure only the output of those sectors (e.g., natural gas, oil) or does it also include the manufactured equipment, supplies, transportation, pipelines and other stuff that necessarily goes into that fracking process that would end up being reported in other sectors?
Aside from that, what would be your guess as to any “multiplier effect”?
31. January 2014 at 12:49
Professor Sumner,
Unrelated, but check out this post by Krugman: http://krugman.blogs.nytimes.com/2014/01/31/the-low-inflationary-trap/?module=BlogPost-Title&version=Blog%20Main&contentCollection=Opinion&action=Click&pgtype=Blogs®ion=Body
He claims that the EU has been at the ZLB for years. Obviously, you have not done enough to dispel this notion.
31. January 2014 at 13:24
Mark,
This is the kind of thing that makes me wonder if “gross value added” sectoral GNP statistics give one a complete picture of the effect that investment in certain sectors have on the “value added” in other sectors and therefore the entire economy:
http://econbrowser.com/archives/2014/01/big-oil-companies-spending-more-and-producing-less
31. January 2014 at 13:40
Mark Sadowski:
“This is hugely debatable, especially since probably the biggest initial response to the crisis by the Fed consisted of nearly $600 in dollar swaps whose main beneficiary was the ECB. So is Trichet supposed to get credit for Bernanke pulling the ECB’s arse out of the fire?”
You are confusing money and credit here. While dollar swaps have eased credit conditions in the Eurozone, they have eased the monetary conditions in the US. By reducing demand for euro denominated ECB facilities, dollar swaps have tightened monetary conditions in the Eurozone. So it was Trichet who has saved Bernanke’s monetary policy by providing dollar liquidity. Dovish Federal Reserve Bank of Frankfurt has saved Bernanke who was surrounded by hawks. Trichet got easier eurozone credit conditions in return. Dollar swaps have made US monetary policy somewhat easier, and they have made euro monetary policy somewhat tighter, but even after this ECB monetary policy remained easier than US as measured by inflation breakevens.
“Also it’s hard to forget that the ECB *raised* the MRO rate in July 2008 which happened to be precisely when the financial markets started blinking red alert warning signals. In fact I am of the opinion that the original impulse for the global great financial crisis started in Frankfurt, not Washington”
It was a wrong move, but Trichet was acting under assumption that Lehman will be saved. The Fed knew the risk better. Do not forget that is wrong to reason from a change in interest rates. Check the EURUSD exchange rate since July 2008 and you will see that Trichet’s policy tightening in the summer of 2008 was much milder than Bernanke’s.
“What possible Taylor Rule are you talking about”
There are studies indicating that 2011 policy action is consistent with the ECB’s prior Taylor Rule reaction function.
31. January 2014 at 15:16
@TravisV,
True, year on year headline PCEPI inflation fell from 2.8% in 2011Q3 to 0.9% in 2013Q4. But core PCEPI only fell from 2.0% in 2012Q1 to 1.1% in 2013Q4 and the GDP price deflator fell from 2.2% in 2011Q3 to 1.4% in 2013Q4. So the decline isn’t really that dramatic when you look at the broadest measure of inflation:
http://research.stlouisfed.org/fred2/graph/?graph_id=157922&category_id=0
@Steve,
Between 2008 and 2013 US oil production increased by 2.65 million barrels per day, which is about 3% of global production. Given the long run price elasticity of demand for oil (the long run in energy is about five years) is about 0.3, that means the price of oil would have been about 10% higher had that increase in production not taken place. That comes to about 0.5% of global NGDP. It’s not hard to see that the benefit of that is both small and would be very widely distributed.
As for natural gas, the elasticity of demand is even higher than oil but the problem is it is difficult to distribute. Thus the effect has been a growing difference in the price of natural gas in North America relative to other continents. According to the World Bank the gap between US and European natural gas prices increased by about $4.16 per mmbtu between 2010 and 2013. Had the gap not changed that would mean the cost of US consumption of natural gas would be about 0.6% of GDP higher relative to Europe. That’s a nice little bonus for US AS but it still does absolutely nothing to explain the huge differences in AD between the US and the Euro Area.
@Vivian Darkbloom,
That’s an interesting question. The most recent detailed input-output data that the BEA makes available is for 2007. What it shows is that the intermediate consumption of the mining and quarrying sector is equal to about 52.0% of that sector’s gross value added or less than 1% of total gross value added. Moreover the proportion of intermediate consumption is less in mining and quarrying than any sector with the sole exception of retail trade where it is 51.4%.
The sector with the highest level of intermediate consumption is manufacturing where it is 199.9% of gross value added. So if such spillovers exist (I’m skeptical) they are far more likely to be in almost every other sector of the economy than mining and quarrying, and would be by far the most significant in manufacturing.
@Vaidas Urba
That’s a very different spin on the dollar swaps than one usually hears in the US, but on first glance it actually makes sense. I’ll have to sleep on that.
As for the MRO change, even if US monetary policy was effectively tightening more than the Euro Area’s in the summer of 2008, an increase in the policy instrument rate is still a much stronger signal of intent than simply standing pat.
“There are studies indicating that 2011 policy action is consistent with the ECB’s prior Taylor Rule reaction function.”
Do you have a link?
31. January 2014 at 15:34
Scott said: “The Fed did poorly over the past 5 1/2 years.”
OK, seems like a straightforward judgment.
Then Scott said: “There was absolutely no way that Bernanke could have implemented the MM agenda even if he had wanted to. Judging the Fed chairman by this criterion would be like judging a President of the US on the assumption that he was a dictator freely able to enact the preferred agenda of a Bryan Caplan or Matt Yglesias.”
I can’t help but wonder if you’re practicing for a second career in politics.
31. January 2014 at 15:45
JC, You said:
He caused the GFC, Scott. You’ve said so in so many words too.”
Never said that. I’ve said the economics profession and the Fed caused the GFC. Big difference.
Everyone, think about the following. Suppose someone else had been elected in 1928 (not Hoover). Say Coolidge. And suppose that because of the policy differences from this counterfactual president the Great Contraction of 1929-33 was only 2/3rds as bad. That person would have been a GREAT president. And yet ALL historians would have said he was horrible. That’s what people miss.
James, You said;
“It almost seems as if the Fed really didn’t like to doing the 2013 monetary offset and can’t wait to return to “normal”.”
I agree.
Dan, A bit slower growth, no big deal.
31. January 2014 at 16:06
Vivian Darkbloom,
This is really weird but just after I left that comment the BEA integrated the 2007 benchmark input-output data into the industry accounts and made even more supplemental estimates (Excel) available (it literally happened as I was navigating the industry section):
http://www.bea.gov/industry/io_annual.htm
I have 2012 data now and it shows that intermediate consumption in the mining and quarrying sector was 34.7% of gross value added in 2012 which is lower than any other sector. Intermediate consumption in manufacturing was 197.8% of gross value added.
Intermediate consumption was $142.5 billion in the mining and quarrying sector in 2012 or about 0.9% of GDP.
You can find it in:
“Use Tables/After Redefinitions/1997-2012: 15 Industries”
31. January 2014 at 17:50
I won’t pull out the hammer because, admittedly, I did start to like Bernanke toward the end of his tenure.
As you said, he did very well in the latter part of 2012 to get support for QE3 and in his March 2013 Congressional testimony. The problem is that I am not sure what he thinks because by June he sounded like the old Bernanke was back. Perhaps in June it was the institutional problems speaking through him, or maybe he got cold feet – I cannot tell. But it probably doesn’t matter so much as he might not have been as isolated as what you describe here if he had exercised some leadership on macro matters from the get go. If he had, it wouldn’t have mattered to people on the extreme fringes of either side of politics, but they are minorities even when combined. It might be naivete on my part, but I think most people would have at least listened to what he had to say, and he never gave them a chance until it was too late and the media spin and a couple of election cycles had already framed of the debate.
Those years Bernanke spent talking about Fed credibility on inflation as a reason for not doing more as NGDP was plunging were lost opportunities to have an impact on what is, if he indeed felt differently than his words let on. But alas, he is only human after all. When it comes down to it, we all are a mixed bag of success and failures, and I can’t rationally expect much more.
31. January 2014 at 18:55
@Mark Sadowski (and Scott).
I was also going where Vivian was going with regards to fracking and its impact on growth during the Great Recession. I remember in late 2008 (even giving the extraordinary monetary policy that was beginning) thinking, “Wouldn’t it be great if we could get lucky with some positive supply shock.” My fantasy was a huge discovery of oil in an unexpected place— my dream was Arizona. Light sweet crude at Saudi cost of production. Spindletop all over again. $1/gal gas…I did not even consider the nat gas component. Obviously fracking is not all that, but its pervasive impact is not likely completely picked up by the Leontieff Input-Output methodology. Think of North Sea Oil years back. The marginal producer can impact all consumers! Energy (nat gas and oil) input costs for, at least all North American users ,have been very favorably impacted. When Scott says that we grew at 2.5% or so and we always have, so there is nothing special about that, I feel that the positive energy situation was about the only thing we had going for us then. The IO model may do a good job of picking up inputs (“Use”), but the upstream effects (in all the users value chains) is probably underestimated.
31. January 2014 at 19:27
“Given the long run price elasticity of demand for oil (the long run in energy is about five years) is about 0.3”
Mark, isn’t the elasticity of demand for oil strongly correlated to the elasticity of supply for natural gas and the political elasticity of supply for coal?
If I’m right then the elasticity of demand for oil would have been lower than historical average in 2008, at least until fracking increased the availability of natural gas.
31. January 2014 at 20:45
Jon Hilsenrath on 9/28/12:
“How Bernanke Pulled the Fed His Way”
http://online.wsj.com/news/articles/SB10000872396390444549204578020252883039778
1. February 2014 at 01:30
Mark,
Thanks for your responses. One can certainly argue over the magnitude, but I guess this confirms what I suspected: those little boxes we like to put things into are not quite as airtight as we sometimes think (that’s one reason we can argue over the magnitude).
I guess input/output data is one way to try to measure; another would be to estimate a private investment multiplier (an issue you didn’t address). My best guess would be that a private investment multiplier would be larger than a fiscal multiplier. You can try to measure an investment multiplier by intermediate consumption, but does that capture ripple effects (that probably extend beyond North Dakota)?
Also, my best guess would be that based on input/output, developing industries like fracking would, in the initial stages, have its effect understated under mere sectoral analysis for the simple reason that before one starts to get stuff out of the ground one has to make significant investments in equipment, etc. Caterpillar’s output doesn’t show up in the mining sector (to my knowledge) and I think you now agree with that. And, in this development stage, the effect seems to be in one direction, or predominately in one direction. Caterpillar’s increased output as a result of this investment doesn’t immediately affect fracking’s output.
Now, you’re the numbers guy and I’m just a relative mathematical illiterate who relies mostly on basic arithmetic. But, despite my inability to attack the problem with advanced differential calculus, I’m skeptical that simply looking at the size of the output of the mining sector to the total economy, or the intermediate inputs to the total economy tells us everything we need to know about the contribution to *incremental growth* to that economy.
But, again, thanks for your responses, which I found edifying.
PS. I trust that based on our exchange here you won’t go over to Econbrowser and announce that you “cleaned the floor with (my) arse”. 🙂 🙂
1. February 2014 at 04:50
Mark Sadowski:
“on first glance it actually makes sense. I’ll have to sleep on that.”
I would be glad to hear your thoughts when you make up your mind.
“an increase in the policy instrument rate is still a much stronger signal of intent than simply standing pat.”
In fact, the phrase “elevated state of some indicators of inflation expectations” in June 2008 FOMC statement is a very strong signal of intent. And it represented a much larger shift in policy.
“Do you have a link?”
I remember seeing a lot of studies, here is just a random link from google:
http://www.europarl.europa.eu/document/activities/cont/201109/20110913ATT26491/20110913ATT26491EN.pdf (chapter 3)
1. February 2014 at 04:52
Mark Sadowski:
Here is an example that uses a form of Taylor Rule that corresponds to ECB’s previous thinking more closely:
http://www.voxeu.org/article/low-how-long-estimating-ecb-s-extended-period-time
1. February 2014 at 07:26
gofx,
“Obviously fracking is not all that, but its pervasive impact is not likely completely picked up by the Leontieff Input-Output methodology.”
The question specifically concerned intermediate consumption which is something which is reasonably easily verified by use of IO Tables.
“Think of North Sea Oil years back. The marginal producer can impact all consumers! Energy (nat gas and oil) input costs for, at least all North American users ,have been very favorably impacted.”
Well, a huge gap has opened up in natural gas prices between North America and other continents because natural gas is rather difficult to transport (and this has happened historically in various parts of the globe).
And yes, West Texas Intermediate (WTI) has been below world oil prices consistently on a monthly basis since November 2010. But the difference has never been more than 15% on a monthly basis and as of last month it was only 7%. And that is the crude oil price, not the price to end users which I’m sure is much smaller on a relative basis.
Since US oil prices became integrated into the world market in 1979 differences of more than 10% between WTI and world oil prics have never persisted for a few months at a time for the simple reason that this would present arbitrage opportunities.
And in the final analysis the gap in natural gas prices that has opened up from 2010 and 2013 between the US and Europe only amounts to about 0.6% of US GDP.
“When Scott says that we grew at 2.5% or so and we always have, so there is nothing special about that, I feel that the positive energy situation was about the only thing we had going for us then.”
This is true provided one totally ignores the huge difference in aggregate demand which even now can not be pumped out of the ground.
“The IO model may do a good job of picking up inputs (“Use”), but the upstream effects (in all the users value chains) is probably underestimated.”
Actually the IO tables work in both directions, so it’s possible to detect how much a part mining and quarrying are of the intermediate consumption of other sectors, and needless to say it is very, very small.
1. February 2014 at 07:29
Steve,
“Mark, isn’t the elasticity of demand for oil strongly correlated to the elasticity of supply for natural gas and the political elasticity of supply for coal?”
Theoretically all three are substitutes for each other, especially in electical power generation and in commercial uses. But as a practical reality the cross price elasticity of oil and natural gas is positive, meaning they are effectively complements. So really we’re talking about to what degree coal can be substituted for oil and natural gas.
“If I’m right then the elasticity of demand for oil would have been lower than historical average in 2008, at least until fracking increased the availability of natural gas.”
The figure I gave for the estimated long run price elasticity of demand for oil is for the world as a whole. The estimated figure for the US (0.4-0.5) is even higher. But given we were talking about the impact of increased US production on the price of oil, and not the impact of a change in the price of oil on US consumption, the relevant figure really is the for the world as a whole, since despite gaps now and then, the price if oil is fundamentally globally determined.
1. February 2014 at 08:37
Vivian Darkbloom,
“I guess input/output data is one way to try to measure; another would be to estimate a private investment multiplier (an issue you didn’t address). My best guess would be that a private investment multiplier would be larger than a fiscal multiplier. You can try to measure an investment multiplier by intermediate consumption, but does that capture ripple effects (that probably extend beyond North Dakota)?”
The differences in fiscal multipliers are largely attributable to the amount of the fiscal policy change that actually results in changes in spending. For example it is typically assumed that tax cuts have smaller multipliers than spending increases (tax cuts are less likely to be spent) or that high income tax cuts have smaller multipliers than low income tax cuts (supposedly high income tax payers have a lower marginal propensity to consume) or that government consumption and investment has a higher multiplier than government transfer payments (by definition, government consumption and investment is 100% spent). With intermediate consumption there is no doubt that spending has taken place, so in my opinion the spending multiplier of a sector should be directly proportional to the ratio of the sum of gross value added and intermediate consumption to gross value added. Assuming this is correct mining and quarrying should have the smallest spending multiplier and manufacturing the largest.
This also matches much of what we know from the economics of development. It’s generally acknowledged that manufacturing has large spillover benefits whereas there is large body of literature devoted to explaining why primary exports (such as oil and natural gas) typically inhibit rather than promote economic development.
“Caterpillar’s output doesn’t show up in the mining sector (to my knowledge) and I think you now agree with that. And, in this development stage, the effect seems to be in one direction, or predominately in one direction. Caterpillar’s increased output as a result of this investment doesn’t immediately affect fracking’s output.”
Caterpillar’s output doesn’t show up in the mining and quarrying sector’s gross value added but it *does show up* in the mining and quarrying sector’s intermediate consumption. But the fact is for the belly aching by US mining and quarrying companies about all the capital spending they must do in order to produce more oil and natural gas in the US they actually spend very little on manufactured equipment used in mining and quarrying (only $38.3 billion in 2012).
“Now, you’re the numbers guy and I’m just a relative mathematical illiterate who relies mostly on basic arithmetic. But, despite my inability to attack the problem with advanced differential calculus, I’m skeptical that simply looking at the size of the output of the mining sector to the total economy, or the intermediate inputs to the total economy tells us everything we need to know about the contribution to *incremental growth* to that economy.”
Well, everything I’ve done up to now has involved no more than basic arithmetic. I’ve actually been reading up on the econometric literature on energy consumption and economic growth and it is actually eye opening for what it does find (or rather, for what it does not). By my count there are 11 different studies on the US alone which look at the relationship between energy consumption and 13 different variables. These studies all use vector autoregression (VAR) and determine causality using Sims, Granger and cointegration tests. Since causality can go in either direction there are a total of 26 test. But out of that 26 there are only six positive results. There are two positive results for GDP causing energy consumption. There is one positive result for employment causing energy consumption. There is one positive result for energy consumption causing employment but the impulse response is negative (more energy consumption results in less employment) and one positive result for energy consumption causing nonfarm employment but again the impulse response is negative. There is only one result for energy consumption causing GDP.
In short, if more energy consumption is the pathway to prosperity, it certainly doesn’t show up in the econometric literature.
Now I realize that energy production is a different question from energy consumption, but presumably we’re currently talking about spillover benefits, and one of the spillovers from producing more energy more cheaply (I would think) is the ability to consume more energy.
“I trust that based on our exchange here you won’t go over to Econbrowser and announce that you “cleaned the floor with (my) arse”.”
To my knowledge I have only ever said that about one person and he had it coming. Kopits every so often bombs the Money Illusion with his deranged comments and then the people with PDS (Petroleum Derangement Syndrome) show up in quick order to to add to the stupidity. Your questions were thoughtful and raised new issues worth pondering over.
1. February 2014 at 08:41
Vaidas Urba,
“I would be glad to hear your thoughts when you make up your mind.”
Still thinking (I’ve been distracted by all of the comments about energy) but I’m inclined to agree with you.
Thanks for the links.
1. February 2014 at 10:28
Mark Sadowski,
I think I speak for everyone here when I say:
You Are Awesome!!!!!
Thanks for all the hard work.
1. February 2014 at 12:37
Sadowski,
What is to be consumed, must be produced. And in energy, almost anything produced will be consumed, right?
Do let’s talk coal. If we cast off all regulations, dismantle EPA, and removes all Oil / Ethanol subsidies….
Will there be more or less energy produced / consumed?
There will be more.
A LOT MORE.
My only real problem with macro guys is that they want to insert themselves into the discussion after the rules and regulations have been established.
The makes them tacit supporters of status quo on rules and regulations.
We all no monetary can be used to bludgeon / smooth over given policies.
So in your 26 studies on the effect of energy of GDP (and again we also know GDP is a horrible measure of digital consumptions) how many of them simply looked at consumption of fuel as the price of fuel falls?
Because we DO consume more when the price of energy falls, right?
I’m just trying to keep the macro literature from disengaging itself from the brutal atomic reality of the petro state.
1. February 2014 at 13:36
Morgan Warstler,
“Because we DO consume more when the price of energy falls, right?”
We also consume more toasters when the price of toasters falls. Yet, so far, nobody claims that increased toaster production will lead to a wave of economic prosperity.
“I’m just trying to keep the macro literature from disengaging itself from the brutal atomic reality of the petro state.”
I’m just trying to keep the petroindustry dogma from relieving itself all over my nice clean commonsense.
1. February 2014 at 13:46
Mark Sadowski. Thanks for your reply. My point was not to analyze the intermediate consumption effects of fracking, but to add to it the knock-on effects which may not be adequately captured in your and Scott’s discsussion. That’s what a supply shock typically is, a great increase in available quantity and/or a great decrease in price (vis-Ã -vis what would have happened without).
IO models have limitations associated with their linearity and the ability to handle new technology. The IMPLAN model is probably a much better IO model than BEA when analyzing energy and there has been some work done in this area on fracking with IMPLAN. Maybe the BEA uses it, I can’t tell. Anyway, fracking is a new technology and I doubt its handled properly yet. Also, I have had more success estimating cost/production functions and impacts using a translog or GL model than an IO. IO is a good “accounting model” and yes it got Leontief his Nobel.
“Well, a huge gap has opened up in natural gas prices between North America and other continents because natural gas is rather difficult to transport (and this has happened historically in various parts of the globe).”
And this is precisely giving a cost advantage to U.S. users of nat gas as a feedstock and fuel source. And its not its % of GDP that matters of this price gap that is relevant, it’s the impact on marginal production and investment decisions. Don’t we talk about expectations on this blog? Incidentally the nat gas is transported abroad, and I am not talking about LNG. Its is transformed into other products (chemicals etc.) which are then sold domestically or exported.
“And yes, West Texas Intermediate (WTI) has been below world oil prices consistently on a monthly basis since November 2010. But the difference has never been more than 15% on a monthly basis and as of last month it was only 7%. And that is the crude oil price, not the price to end users which I’m sure is much smaller on a relative basis.”
7%-14% price differentials are important to firms who use a lot of petroleum derived products. For domestic U.S. production, I disagree that the end-use cost differential is much lower.
Here you argue that for oil, it’s a world market anyway so there is no domestic advantage as in gas. But this too is precisely my point and the North Sea reference, fracking has impacted the world price of oil!
“This is true provided one totally ignores the huge difference in aggregate demand which even now can not be pumped out of the ground.”
This is really the crux. My point is that given that you and Scott think that the Fed underperformed in this period, and that it seem like they were operating more or less on an inflation (and employment) target, not NGDP, I am saying that the positive supply shock in U.S. domestic oil and gas production was more significant than you think. Try this thought experiment (and remember that expectations are not just about monetary policy, but also input prices): Keep the Fed actions constant (since inflation constraint was nonbinding) and tell me what you think AD would look like with crude at say $10/barrel vs. $70-$100. Maybe AD is just a little higher? And for grins, now does the same thought experiment, with appropriate prices, for a revolution in the production of say, breath mints.
1. February 2014 at 14:00
Mark Sadowski
BTW. I do not, and have never worked for an oil or nat gas company. I do drive a car and have worked for companies that consume oil and natural gas. I also consume breath mints.
1. February 2014 at 14:20
“We also consume more toasters when the price of toasters falls. Yet, so far, nobody claims that increased toaster production will lead to a wave of economic prosperity.”
The reason for the difference is obvious. Expenditures on toasters are a miniscule percentage of total expenditures and are also miniscule input in the production of goods on which there are a lot of expenditures. Therefore the price of toasters are one of the things that can be assumed to remain equal when looking at factors that produce economic prosperity.
1. February 2014 at 15:06
“So Goolsbee views this as success, but never tells us why. He never explains what the Fed should be targeting, or by what criterion he judged the policy a success.”
There is no objective way to judge the success of shooting at innocent people. Of course you will always have disagreement among those wedded to irrationalism.
Same thing with legalized, gun-backed counterfeiting. There is no objective way to measure the success of harming innocent people.
You will never find an intellectual justification for MM theory, because MM theory is based on violence, not reason.
1. February 2014 at 15:08
Clearly petroleum is a major input in production and expenditures on petroleum products are a significant part of total expenditures so the effects of petroleum production and prices are not one of the things that can be assumed to remain constant. So the question is how important an effect did this have?
1. February 2014 at 15:39
“Same thing with legalized, gun-backed counterfeiting. There is no objective way to measure the success of harming innocent people. You will never find an intellectual justification for MM theory, because MM theory is based on violence, not reason.”
Talk about irrationalism. This dogma is as irrational as it gets.
The government or a central bank set up by the government, engaged in conducting monetary policy, involves less violence than the government’s protecting the property rights of the rich. Less wealthy people, if they try to take control of some of this property by force, are subject to forcible arrest, and not infrequently end up being killed by the police. If use of violence to uphold property rights (and contacts) is justified the use of violence to uphold a regime of monetary policy is similarly justified.
Economic theories that call for the use of monetary policy to achieve full employment are much more rationally developed than the Austrian dogma, which rejects all empirical research, for which there is no proof, and which is held on the basis of faith by its acolytes.
The right of government to coin money and regulate the value thereof became established at the time that rulers put their stamp on lumps of valuable metal to guarantee their quality and weight, and has been established since then.
The use of monetary policy to achieve full employment without any significant inflation helps a lot more people than it harms.
1. February 2014 at 15:45
“There is no objective way to judge the success of shooting at innocent people.”
Yes there is: Society needs to assign value judgements on the social cost of having people killed, compared to having people injured, compared to having people frightened. Then one only has to count the number of people killed, the number of people injured, and the number of people frightened and use those value judgement to assign weights to the number of each category and add the rusults up.
1. February 2014 at 16:59
gofx,
“IO models have limitations associated with their linearity and the ability to handle new technology. The IMPLAN model is probably a much better IO model than BEA when analyzing energy and there has been some work done in this area on fracking with IMPLAN. Maybe the BEA uses it, I can’t tell. Anyway, fracking is a new technology and I doubt its handled properly yet. Also, I have had more success estimating cost/production functions and impacts using a translog or GL model than an IO. IO is a good “accounting model” and yes it got Leontief his Nobel.”
No offense intended but this reveals deep confusion.
The BEA’s IO tables do not come from a model. They come from a variety of sources including the Census Bureau, the Departments of Agriculture, Education and Energy, and a number of private sources. The BEA’s IO tables are used to compute the NIPA accounts including the nation’s GDP.
Moreover they are source data for the Department of Commerce’s RIMS II Model, the IMPLAN Model and the even more sophisticated REMI Model. So rather than those models being better than the BEA’s “model”, the BEA’s IO tables are literally the benchmark from which those models are constructed.
“And its not its % of GDP that matters of this price gap that is relevant, it’s the impact on marginal production and investment decisions. Don’t we talk about expectations on this blog?”
The impact of the differential is on the entire aggregate supply curve so it *is* precisely the % of GDP that matters from a macroeconomic perspective. Furthermore speaking from the standpoint of rational expectations I don’t expect this differential to persist for too long.
“7%-14% price differentials are important to firms who use a lot of petroleum derived products. For domestic U.S. production, I disagree that the end-use cost differential is much lower.”
The differential averaged 5.9% in 2013 and generally has trended downward since reaching its peak in September 2011. Nearly half (46.4% in 2013) of all US oil consumption in the US is in the form of gasoline, and as of 2010 (the most recent estimates available) crude oil represented 68.9% of the cost of a gallon of gasoline.
“Here you argue that for oil, it’s a world market anyway so there is no domestic advantage as in gas. But this too is precisely my point and the North Sea reference, fracking has impacted the world price of oil!”
But if fracking impacts global price of oil than how is the US economy to extract a unique benefit from this aspect of fracking other than the gross value of the oil production it adds to the US economy?
“Try this thought experiment (and remember that expectations are not just about monetary policy, but also input prices): Keep the Fed actions constant (since inflation constraint was nonbinding) and tell me what you think AD would look like with crude at say $10/barrel vs. $70-$100. Maybe AD is just a little higher?”
In my opinion, there is considerable evidence that the FOMC has recently been behaving as though it is targeting a 1.6% core inflation rate. Were oil to drop to $10 a barrel that would be the equivalent of a huge shift in the short run AS curve to the right, but I don’t think that would push down core inflation by very much. So yes this would temporarily result in big increase in the rate of growth in RGDP but I don’t think the rate of change in NGDP would increase by very much.
1. February 2014 at 17:51
Full Employment Hawk,
“The reason for the difference is obvious. Expenditures on toasters are a miniscule percentage of total expenditures and are also miniscule input in the production of goods on which there are a lot of expenditures. Therefore the price of toasters are one of the things that can be assumed to remain equal when looking at factors that produce economic prosperity.”
You have succeeded in missing the entire point of that analogy by a country mile.
The gross value added by the entire mining and quarrying sector, of which oil and natural gas extraction is a subset, was only 1.75% of total gross value added in 2013Q3:
http://research.stlouisfed.org/fred2/graph/?graph_id=141978&category_id=0
How much does the gross value of toaster manufacture contribute to the US economy? There’s no way of knowing precisely, but appliances, electrical equipment and components, of which toasters are a subset, represent about 0.3% of total gross value added. In short, the two sectors are of the same magnitude.
1. February 2014 at 21:11
Mark Sadowski.
There is no confusion. I simply think that the IO is unlikely to represent a newer technology accurately. But maybe it will/does. Of course IO is a model! It’s a general equilibrium-inspired linear, static, model that assumes NO economies of scale, and that each industry has a unique production function, no alternate use of slack resources, coefficients are fixed, and at least in the past, the law of one price holds in an industry, and prices don’t change. Even Leontief recognized this, and still argued that his general empirical approach was better than what I am advocating (a more specific analysis). And if you think then, its just a “NIPA accounting table” then it sounds a lot like MMT’ers worshiping accounting identities. What is everybody going to do (unless funding got restored) now that RIMS II multipliers are going away?
On the nat gas and oil impacts accruing to the U.S. I was responding your point on the global market for oil (with which I agree). I am fine with a unique advantage to the U.S. on nat gas and a prorata share of the benefit of a reduction in the global price of oil. If there is any unique effect to the U.S on local oil production, that’s a bonus. Even if the nat gas was a world market, the increase in supply is beneficial.
1. February 2014 at 22:45
While hesitating to enter into a debate with Sadowski, who is leagues above me in terms of economic understanding, isn’t the best way to look at the impact of shale gas and oil on the terms of trade? Daniel Yergin just published an estimate (http://peakoil.com/production/daniel-yergin-the-global-impact-of-us-shale) where he states that the combined impact on reduced imports of LNG and oil is worth $200bn per month of reduction in imports. This means an annual $2.4trn of payments that instead of going overseas, stays in the US. Eventually this money must find its way to consumers pockets adding to US consumption. By contrast, the 2009 ARA stimulus package was estimated to $831bn between 2009 and 2019.
1. February 2014 at 22:47
Oops, my mistake, the figure quoted by Yergin was $200bn annual not monthly. I think the point still stands, certainly the impact is similar in magnitude to the ARA
1. February 2014 at 23:02
ChrisA,
Yergin, via IHS also estimates current fracking value to have been worth $1200 per U.S. Household (“American’s New Energy Future”). They used a modified IMPLAN model.
PwC has a report “Shale Oil: The Next Energy Revolution 2013” where they used the NiGEM econometric model to estimate the macroeconomic impacts of the shale impact. Model documentation says it s NK model.
Since fracking has become such a political hot potato (environment) its hard to know what to make of many of these studies.
2. February 2014 at 07:41
gofx,
“There is no confusion.”
You said:
“The IMPLAN model is probably a much better IO model than BEA when analyzing energy and there has been some work done in this area on fracking with IMPLAN. Maybe the BEA uses it, I can’t tell.”
Why on earth would the BEA use IMPLAN to estimate anything when, at least in benchmark years, the BEA has virtually all the raw data?!?
“I simply think that the IO is unlikely to represent a newer technology accurately. But maybe it will/does.”
IO tables have been used by the BEA since 1958, and by the countries using SNA since 1953. If there was a huge problem representing newer technologies in IO, don’t you think it would have been discovered before now?
“Of course IO is a model!”
If the BEA’s IO tables are a model then at the very least it is a survey based model. I guess it’s a question of what one considers to be a model. I personally don’t consider a table of values largely collected by survey to be a model.
“It’s a general equilibrium-inspired linear, static, model that assumes NO economies of scale, and that each industry has a unique production function, no alternate use of slack resources, coefficients are fixed, and at least in the past, the law of one price holds in an industry, and prices don’t change.”
It’s a table of annual values, so of course it is static. Economies of scale only matters for values which muct be estimated, which in benchmark years are practically nonexistent. The assumption of homogeneous production functions is only relevant to the manner in which the data is disaggregated. This has no impact at all on the accuracy of the data itself, only on models that use the data. As for the rest of your concerns, where do you think the estimates of the parameters come from?
“Even Leontief recognized this, and still argued that his general empirical approach was better than what I am advocating (a more specific analysis).”
Leontief was an empiricist. Collecting raw data may be time consuming and tedious but it trumps theoretical assumptions every time.
“And if you think then, its just a “NIPA accounting table” then it sounds a lot like MMT’ers worshiping accounting identities.”
The problem with MMT is that they treat accounting identities as behavioral functions. But at least they don’t confuse accounting identities with models.
“What is everybody going to do (unless funding got restored) now that RIMS II multipliers are going away?”
Regional, state and local impact estimates are going to become far less reliable and much more expensive to produce. But at least the Republicans have succeeded reducing the deficit by (gasp) $1.4 million dollars:
http://scifundchallenge.org/wp-content/blogs.dir/28/files/2013/06/Dr_-Evil-One-Million-Dollars_zps107ab072.png
However, on the bright side, this will not affect the BEA and its estimates of NIPA and GDP one bit.
2. February 2014 at 08:25
ChrisA,
“…isn’t the best way to look at the impact of shale gas and oil on the terms of trade? Daniel Yergin just published an estimate where he states that the combined impact on reduced imports of LNG and oil is worth $200bn per [year] of reduction in imports…”
Yergin appears to be assuming that the trade deficit will fall by the same amount that shale oil and natural gas reduces US imports of petroleum products. Is that a reasonable assumption?
The answer is of course no. Theoretically the trade balance of a given country is largely a function of its savings and investment behavior. Importing less petroleum products is unlikely to alter that behavior much if at all.
Moreover this is not something we even have to depend on economic theory to know, we have empirical evidence. For some of that I’ll turn the microphone over to Robert Z. Lawrence:
“…Looking at the historical record can confirm some of these insights. Separating out the precise influence of oil on the current account in historical data from other sources of influence is challenging, but it is still possible to learn from past experience. The correlation between the aggregate movement in the current account and the net oil trade balance has been weak. As shown in Figure 3, two major oil price increases in 1973 and 1979 led to large deficits in net oil trade, but these shocks were offset by improvements in the non-oil current account””in part because they led to recessions. Between 1981 and 1987, by contrast, a shrinking oil deficit coincided with a dramatic decline in the current account (i.e., increasing the current account deficit) because non-oil trade moved in the opposite direction due to the strong dollar brought about by the high interest rates associated with Reaganomics. Although the movement in the two deficits was more closely correlated between 1999 and 2005, the chart clearly suggests that the dominant determinants of the current account lay elsewhere””an account pointing to the importance of taking aggregate saving and investment decisions into account.
More powerful evidence of the weak long-run relationship between oil deficits and current account deficits internationally is provided by Joseph Gagnon in a 2013 study.14 In his study, which uses two samples of 40 and 115 countries, Gagnon is actually interested in explaining the role of currency intervention in driving the current account. But his results are useful for present purposes because among the many variables he uses in his regressions to explain the current account is each country’s net energy-trade balance. Thus, one can interpret his regressions as indicating the effect of changes in the energy trade balance while controlling for other relevant independent variables.
Gagnon’s work is thorough. His regressions explain the annual and five-year behavior of the countries’ current accounts between 1986 and 2010. He presents a large number of specifications and takes great care to deal with potential statistical problems such as simultaneity bias. His results confirm that changes in the net energy trade balance have relatively small effects on the current account, especially over the longer run. When he uses five-year country averages for current account, he typically finds that for each dollar improvement in the energy balance, the current account improves by just ten cents. When using annual data, the improvement is typically twice as large.15 Another statistical approach involves the use of vector-autoregression models to explore the effect of oil price shocks. In one such exercise, the International Monetary Fund finds that permanent “oil price shocks have a marked but relatively short-lived impact on current accounts,” reinforcing Gagnon’s analysis and the broader arguments presented here.16…”
You can find that passage on pages 10-11 of his January 2014 CFR Energy Report which can be downloaded here:
http://www.cfr.org/united-states/implications-reduced-oil-imports-us-trade-deficit/p32245
I also highly recommend the Joseph Gagnon paper to which he refers:
http://www.iie.com/publications/wp/wp13-2.pdf
2. February 2014 at 12:44
Mark Sadowski
I think you hit the nail on the head, we are in disagreement about what constitutes a model (I don’t believe you have to have an epsilon to have a “model”, and a model may be implicit.)
I’m perplexed, you seem to say the BEA data are just NIPA accounting identities, but then you are trying to assess the impact of a technology change with identities, much as the MMTers are doing with monetary (which you rightly criticize). Conversely if there is an implicit or explicit model underlying BEA/IO then such an underlying set of assumptions should be allowed to be questioned as to conclusions about impacts of large technological changes.
Behaviors can change even while you are looking at your static IO tables! A funny thing happened on the way through my IO table–somebody changed their behavior! The double-entry bookkeeping system has an underlying model, FASB creates the accounts and rules, but that does not necessarily capture the all economic implications of changes in a firms environment.
“This has no impact at all on the accuracy of the data”
Its not the “accuracy” of the data, its what the data represent in the context of trying to figure out an impact of a technological change.
“IO tables have been used by the BEA since 1958, and by the countries using SNA since 1953. If there was a huge problem representing newer technologies in IO, don’t you think it would have been discovered before now? ”
Well, yes there have.
First, some of the more detailed benchmarks from BEA are only updated every five years if I recall. Not exactly what we’d like.
Second, that’s why some people approach this type of research econometrically. (And yes, per your response to ChrisA, there can be problems there too.)
Third, modelers will modify parts of the BEA data when they have something better or things don’t make sense.
And fourth, that is also why Computable General Equilibrium (CGE) models were developed. Here is Wikipedia:
“CGE models are descended from the input-output models pioneered by Wassily Leontief, but assign a more important role to prices. Thus, where Leontief assumed that, say, a fixed amount of labour was required to produce a ton of iron, a CGE model would normally allow wage levels to (negatively) affect labour demands………
CGE models are useful whenever we wish to estimate the effect of changes in one part of the economy upon the rest. For example, a tax on flour might affect bread prices, the CPI, and hence perhaps wages and employment. They have been used widely to analyse trade policy. More recently, CGE has been a popular way to estimate the economic effects of measures to reduce greenhouse gas emissions.”
Note the use of the term, “model”
Here’s a CGE-based model for fracking impacts and nat gas export impacts using a model called GTAP-E.
http://www.slideshare.net/MarcellusDN/purdue-university-research-economic-impacts-of-shale-oilgas-in-us
Imagine these fools’ surprise when we tell them all they had to do was look at the BEA tables and just stop right there.
Here’s another energy one coming down the line: energy storage. People are working on breakthroughs in capacitors that would greatly increase electricity storage (and better than batteries), let’s say by 100%. I just don’t see the impact jumping off of the IO table.
But I have to note here, I have not done any modeling to assess the fracking impact.
2. February 2014 at 17:02
gofx,
“I’m perplexed, you seem to say the BEA data are just NIPA accounting identities, but then you are trying to assess the impact of a technology change with identities, much as the MMTers are doing with monetary (which you rightly criticize).”
I criticize the MMTer’s interpretation of the data, not the data itself. You evidently are questioning the data.
“Conversely if there is an implicit or explicit model underlying BEA/IO then such an underlying set of assumptions should be allowed to be questioned as to conclusions about impacts of large technological changes.”
So far the only substantive questions you have raised concern forecasts based on the data, not the data itself.
“Behaviors can change even while you are looking at your static IO tables! A funny thing happened on the way through my IO table-somebody changed their behavior! The double-entry bookkeeping system has an underlying model, FASB creates the accounts and rules, but that does not necessarily capture the all economic implications of changes in a firms environment.”
How does behavioral change make gross value added and intermediate consumption not show up in the BEA IO tables?
“Its not the “accuracy” of the data, its what the data represent in the context of trying to figure out an impact of a technological change.”
What are the BEA IO tables supposed to represent other than gross value added and intermediate consumption?
“First, some of the more detailed benchmarks from BEA are only updated every five years if I recall. Not exactly what we’d like.”
It would be nice if it were possible to do benchmark surveys more frequently than every five years. But has this been a major problem in the past? Not really. (Incidentally the most recent benchmark was in 2012.)
“Second, that’s why some people approach this type of research econometrically. (And yes, per your response to ChrisA, there can be problems there too.)”
Econometrics can be used, among other things, to estimate that which we do not know from direct observation. In this context that typically means forecasts. But there is no need for econometrics when you already have direct observations.
“Note the use of the term, “model”.”
Note the fact that CGE models are useful whenever we wish to estimate the effect of *changes in one part of the economy upon the rest*. In contrast the BEA’s IO tables are snapshots of the economy *as it is*.
“Here’s a CGE-based model for fracking impacts and nat gas export impacts using a model called GTAP-E…Imagine these fools’ surprise when we tell them all they had to do was look at the BEA tables and just stop right there.”
Well, for one thing they are forecasting over 20 years into the future, whereas I am talking about what has happened in the last five years. For another thing, in my opinion their forecasts are wildly implausible. Have there been any CGE forecasts of the economic impact of oil shale that aren’t so ridiculously optimistic?
2. February 2014 at 17:48
Mark Sadowski,
“Well, for one thing they are forecasting over 20 years into the future, whereas I am talking about what has happened in the last five years. For another thing, in my opinion their forecasts are wildly implausible. Have there been any CGE forecasts of the economic impact of oil shale that aren’t so ridiculously optimistic?”
Even the attribution of the impact of something like fracking over the past years typically involves a forecast of the counterfactual. As we say, “forecasting is difficult, especially when its about the future, but the past is pretty hard too”. I am sure there are environmental groups with runs of models that purport to show fracking has a negative impact on GDP or at least net national welfare. I agree with you, though, I think their forecasts look too optimistic.
“Note the fact that CGE models are useful whenever we wish to estimate the effect of *changes in one part of the economy upon the rest*. In contrast the BEA’s IO tables are snapshots of the economy *as it is*.
But our research question is the effect of changes in one part of the economy (fracking) on the rest (GDP)! To truly determine what the effect of fracking has been we need a “snapshot” of the economy “as it would have been” (absent fracking) not “as it is”. And this is precisely why a model is needed (price impacts, substitutions, pattern changes, etc.). Just applying multipliers based on “what has been” to quantity/proportion changes doesn’t get it, in my opinion, but I guess we’ll just have to agree to disagree.
By the way this debate is purely technical. I enjoy and have benefited from your posts and empirical work you put on this and other sites, especially your Granger causality testing.
2. February 2014 at 19:04
Mark, in one of the links you gave I find this quote;
“I show in this paper that a decline in net imports of oil and energy-intensive manufactured goods is likely to be offset by greater net imports in other goods and services. In the long run, the changes in oil and non-oil trade balances could well cancel each other, leading to little or no change in the overall U.S. trade deficit. ”
In other words the higher domestic oil production allowed the US to buy more stuff from overseas, which caused the trade deficit to go back to what it was before the increase. So comparing where the US was before the increase and is now, the net effect is that people have consumed more, by the same amount as the reduced oil imports. In Yergin’s analysis, this is $200bn a year net increase in consumption of goods and services by Americans, surely that would have a positive impact on GDP, even if it were all imported. (Which also begs the question, if there really was an output gap in 2009 through 2013, why didn’t the increase in goods and services happen in the US rather than overseas).
3. February 2014 at 05:59
gofx,
“Even the attribution of the impact of something like fracking over the past years typically involves a forecast of the counterfactual. As we say, “forecasting is difficult, especially when its about the future, but the past is pretty hard too”…But our research question is the effect of changes in one part of the economy (fracking) on the rest (GDP)! To truly determine what the effect of fracking has been we need a “snapshot” of the economy “as it would have been” (absent fracking) not “as it is”.”
In 2012 (a benchmark year) oil and natural gas extraction (by itself) was 1.49% of GDP (gross value added plus taxes on production and imports less subsidies). The intermediate consumption of oil and natural gas extraction was 0.49% of GDP. In the absence of the oil shale boom these quantities would of course be less. So with respect to the direct effect on GDP, the counterfactual is a proportion of some very small quantities.
Now, there has also been an effect on the price of energy. But with respect to oil this has led to a relatively small differential compared to the rest of the world (roughly 4% on average to an end use consumer as of 2013). And with respect to natural gas, although the price differential is relatively large, the total effect is perhaps 0.6% of GDP as of 2013.
I don’t think a more precise estimate of the price impacts, substitutions, pattern changes etc. is going to change the magnitude of these effects.
“By the way this debate is purely technical. I enjoy and have benefited from your posts and empirical work you put on this and other sites, especially your Granger causality testing.”
Thank you very much.
3. February 2014 at 06:43
ChrisA,
“So comparing where the US was before the increase and is now, the net effect is that people have consumed more, by the same amount as the reduced oil imports.”
The total effect is that the US consumes relatively more domestic and less foreign petroleum products, and relatively less domestic and more foreign nonpetroleum goods and services. The net effect on the trade deficit is thus relatively small.
“In Yergin’s analysis, this is $200bn a year net increase in consumption of goods and services by Americans, surely that would have a positive impact on GDP, even if it were all imported.”
Well, combining Yergin’s $200 billion annual estimate of the effect of shale oil and natural gas on the energy balance, with Gagnon’s 5-year country estimate that for each dollar improvement in the energy balance the current account improves by just ten cents, net exports would have increased by about $20 billion a year or about 0.12% of GDP. So the trade effect of shale oil and natural gas on GDP is probably quite small.
“Which also begs the question, if there really was an output gap in 2009 through 2013, why didn’t the increase in goods and services happen in the US rather than overseas.”
Well yes in fact, that is a very good point. A careful reading of Robert Z. Lawrence’s analysis implies that the effect of changes in the energy balance on net exports may be larger when there is an output gap. But in the absence of further analysis, how much larger is an open question.
3. February 2014 at 16:17
Mark
Your focus is on the trade deficit re-balancing or not. But that is not my point, my point is that the consumption of “real” non-oil goods and services by domestic consumers will have to have increased by the same amount as the reduced imports of oil to keep the trade in balance. You seem to me to agree on this point. So with reference to the original question, of whether the shale revolution in the last five years has had a material effect on the economy, the answer must be yes, as it has increased “real” demand in the domestic economy by $200bn a year. Now you could argue that in previous examples the demand is satisfied fully by imports so it didn’t matter to GDP growth, but I am not sure why oil driven demand is different from any other demand increase. So why is this necessarily so in this particular instance? It seems to me that the default assumption should be that when an economy is below potential any domestic demand increase should be filled by domestic supply. That is certainly the assumption being used by advocates of fiscal and monetary stimulus such as the author of this blog.
To put it simply, I still think an extra $200bn per year increase in domestic demand can materially affect domestic growth in a depressed economy.
3. February 2014 at 22:13
ChrisA and Mark,
As I mentioned in one of my earlier comments, Yergin also published some results via IHS. I don’t know if it’s the same as the balance-of-payments results you are discussing, but I was finally able to find some the documentation! (They reference other appendices in the document).
http://www.energyxxi.org/sites/default/files/pdf/americas_new_energy_future-unconventional_oil_and_gas.pdf
It’s a pretty thorough undertaking and maybe there is something here for everybody! Per my dreams, they did in fact use IMPLAN, modified some of the IO relationships to fit fracking supply chains, and then linked it with a macro model and iteratively solved between IMPLAN and the Macro model for economic contribution. Mark, I swear I do not work for them!
They came up with $237B value added to GDP for 2012. It seems high to me, but there was a lot of up front domestic capital investment in all these startups. Have at it!
4. February 2014 at 00:19
Gofx, I read the report and I think it is credible, at least in order of magnitude to the ARA stimulus. I do subscribe to the Sumner critique of course, the Fed basically determines the ultimate level of demand in an economy, by increasing or decreasing money supply offsetting any stimulus whether natural (such as the unconventional or shale industry efforts) or “unnatural” (such as fiscal stimulus). So ultimately, to draw all this back to the original post, the Fed are the reason the economy is growing now, because they did QE. But without the shale they would have had to do more QE, like they are doing in the UK.
Anyway, enough from me. I am playing above my league.
4. February 2014 at 04:44
ChrisA,
“Your focus is on the trade deficit re-balancing or not.”
That’s probably because it was you who suggested that the best way to look at the impact of shale oil and natural was in terms of international trade.
“But that is not my point, my point is that the consumption of “real” non-oil goods and services by domestic consumers will have to have increased by the same amount as the reduced imports of oil to keep the trade in balance. You seem to me to agree on this point.”
But keep in mind that this increase in consumption of foreign non-petroleum goods and services is more or less balanced by a decrease in the consumption of domestic non-petroleum goods and services.
“So with reference to the original question, of whether the shale revolution in the last five years has had a material effect on the economy, the answer must be yes, as it has increased “real” demand in the domestic economy by $200bn a year.”
The shale oil and natural gas boom is effectively a positive shock to aggregate supply (AS). It has no effect at all on the level of aggregate demand (AD). Because it has led to changes in the relative cost of production this is causing demand to reallocate away from foreign oil and natural gas to domestic and from domestic non-petroleum goods and services to foreign.
“Now you could argue that in previous examples the demand is satisfied fully by imports so it didn’t matter to GDP growth, but I am not sure why oil driven demand is different from any other demand increase. So why is this necessarily so in this particular instance?”
It’s a reallocation of demand driven by an AS shock. Moreover this is not unique to energy. Since trade balances are largely a function of savings and investment behavior, any AS shock that causes a country to substitute domestic production for imports of one or more goods and services will exhibit a reciprocal effect where it substitutes imports of other goods or services for domestic production.
“It seems to me that the default assumption should be that when an economy is below potential any domestic demand increase should be filled by domestic supply. That is certainly the assumption being used by advocates of fiscal and monetary stimulus such as the author of this blog.”
Actually this is not at all true. The effect of an increase in AD on the trade balance depends on a number of factors, and even if there is an output gap, empirically it is more likely to decrease net exports than increase them. See in particular the case of Japan since Abenomics where although there has been a huge increase in nominal exports, there has been an even larger increase in nominal imports. In fact Japan just set a new record for its trade deficit.
I also think your generalization of the views of advocates of fiscal and monetary stimulus on this matter is highly questionable. I know for example that extreme fiscalists, like those at the Levy Institute, are quite conscious of the fact that fiscal stimulus can “leak” into decreased trade balances. And as for Scott Sumner, I’m confident he is aware that monetary stimulus often results in decreased trade balances even when there is a large output gap.
“To put it simply, I still think an extra $200bn per year increase in domestic demand can materially affect domestic growth in a depressed economy.”
The problem is it’s *not* increased demand. It is a shift in aggregate supply that manifests itself in a reallocation of demand.
4. February 2014 at 07:23
gofx,
“As I mentioned in one of my earlier comments, Yergin also published some results via IHS. I don’t know if it’s the same as the balance-of-payments results you are discussing, but I was finally able to find some the documentation! (They reference other appendices in the document).”
Thanks. Although it is very long it is also highly repetitive, so I think I’ve already absorbed the gist of it. No, it is not focused on the international trade issue, but at least I can see where Yergin is coming from.
“They came up with $237B value added to GDP for 2012. It seems high to me, but there was a lot of up front domestic capital investment in all these startups. Have at it!”
The estimated effects come in four major categories: 1) capital expenditures, 2) gross value added, 3) employment and labor income and 4) government revenue. The effects on capital expenditures, although large, appear to be confined to the oil and natural gas extraction, but the effects on the other three categories include not only direct effects but indirect effects and “induced” effects. Moreover it’s important to keep in mind that labor income and government revenue are already part of total value added (gross value added plus taxes on production and imports less subsidies), so there is a lot of implicit double counting going on.
Let’s take a look at some of these estimates and see how realistic they are.
1) Capital expenditures
On page 12 of Appendix A it shows that the capital expenditures on “unconventional activities” totaled $87.3 billion in 2012. This compares to the the BEA figure of $80.1 billion and $24.6 billion spent on “intermediate consumption” by the “oil and gas extraction” (NAICS 211) and “support activities for mining” (NAICS 213) respectively. Of the total for capital expenditures, $9.2 billion was spent on “support activities”, leaving $78.1 billion. This suggests that 97.5% of all intermediate consumption by oil and gas extraction and that 37.4% of all intermediate consumption by support activites for mining consisted of capital expenditures on unconventional activities. To see how believable that is we must turn to gross value added.
2) Gross value added
The figures for gross value added are divided into direct, indirect and induced effects on page 31. Direct effects in 2012 are $96.7 billion, indirect effects are $67.2 billion and induced effects are 73.8 billion for a total of $237.7 billion.
It’s not clear how this is allocated by sector but let’s turn to oil and gas extraction on page 10 of Appendix C to get an idea. Here we see that $67.7 billion of this effect is in oil and gas extraction. According to the BEA, there was a total of $193.1 billion in gross value added in oil and natural gas extraction in 2012. Thus 35.0% of all gross value added in oil and gas extraction was in unconventional activities in 2012. Furthermore, it’s safe to assume that this accounts for 70.0% of the direct effects of unconventional activities.
Turning to pages 11 and 12 of Appendix C we see that $4.8 billion and $11.5 billion of this effect is in “mining, except oil and gas” (NAICS 212) and support activities for mining respectively. According to the BEA the gross value added of other mining and support activities for mining was $64.6 billion and $110.9 billion in 2012. Thus 7.4% and 10.4% of the gross value added in other mining and support activities for mining are attributed to unconventional activities. I would assume the portion allocated to support activities is a direct effect. This may also be true of other mining but it’s not clear to me why unconventional activities in oil and natural gas should be stimulating coal and other non-oil and gas mining.
Recall also that 37.4% of all intermediate consumption by support activites for mining consisted of capital expenditures on unconventional activities. This is substantially larger than the 10.4% proportion of gross value added by support activities for mining attributed to unconventional activities.
3) Employment and labor income
On page 27 we learn that in 2012 the direct employment effect was about 360,500 jobs, the indirect effect was 537,700 jobs and the induced effect was about 850,500 jobs for a total of 1,748,600 jobs. On page 33 we learn that these jobs generated $43.6 billion, $39.3 billion and $41.7 billion in labor income respectively for a total of $124.5 billion in labor income.
On page 10 of Appendix C we learn that 111,100 of these jobs were in oil and gas extraction and generated $24.3 billion in labor income in 2012. According to the BLS there were 186,800 people employed in oil and gas extraction in 2012 implying 59.5% of these were employed in unconventional activities. According to the BEA compensation of employees in oil extraction totaled $11.7 billion in 2012. This means Yergin’s estimate of labor income for just those employed in unconventional activities in oil and gas extraction is over double the entire compensation of employees for oil and natural gas extraction. Needless to say that is a huge discrepancy.
Turning to pages 11 and 12 in Appendix C we learn that 35,000 and 49,000 of these jobs were in other mining and support activities for mining respectively in 2012. The labor income of these employees totaled $3.3 billion and $4.6 billion respectively. According to the BLS there were 222,700 and 390,700 employed in other mining and support activities for mining respectively in 2012. According to the BEA the compensation of employees totaled $10.7 billion and $57.6 billion respectively in 2012. Thus 15.7% and 12.5% of the jobs in other mining and support activities for mining respectively are attributable to unconventional activities and 30.8% and 8.0% of the compensation of employees in other mining and support activities for mining respectively are attributable to unconventional activities.
The total employment attributable to unconventional activities in these three subsectors is 195,100. This is substantially less than the 360,500 jobs directly attributable to unconventional activities, so one is left wondering how broad the definition of “direct effect” is.
Moreover where are all the other jobs and labor income such that it adds up to 1,750,000 jobs?
Well on page 73 of Appendix C for example we learn that 141,200 jobs are created in professional, scientific and technical services alone for example. Presumably these, along with 7,300 jobs in the postal service, 7,900 jobs in publishing, 21,000 jobs in central banking, and 14,000 jobs in the performing arts are part of the 850,000 jobs created under the “induced” effects.
4) Government revenue
On page 35 we learn that unconventional activities are responsible for increasing government revenue by $61.8 billion in 2012. This is evenly divided beteen the federal and state and local governments. Of this $25.6 billion is in personal taxes, $26.0 billion is in corporate taxes, and $10.2 billion is in other kinds of taxes.
Up to date IRS SOI Tax Stat records are not really available, but needless to say these revenue figures are only believable if you buy the indirect and induced effects which I do not. The proportion of revenue directly attributable to oil and gas extraction is similar to the tiny proportion of gross value added it generates each year. For example out of the nearly $298.7 billion in federal corporate taxes paid in calendar year 2010 only $5.0 billion were paid by mining and quarrying corporations, of which oil and gas extraction is a subset, and of which unconventional activities is a still smaller subset.
In any case, keep in mind government revenue and labor income are already contained within total value added, so this is just implicit double counting.
All in all this makes for truly mindbending reading. Only someone really enjoys brainwashing themselves would believe everything contained in this report.
4. February 2014 at 13:55
ChrisA,
I don’t think so, you might just be in a league of your own! I agree with your point on the Sumner critique. If the Fed has not hit their true inflation constraint, then maybe NGDP can be favorably impacted, if as Mark has indicated, they have an unarticulated constraint at around 1.6% inflation, then the favorable supply shock will only increase the RGDP share of NGDP.
4. February 2014 at 14:26
Mark,
Well you’re right, I am not going to bend my mind and go through that report in detail beyond reading the basic report. Kudos to you for doing as much as you did. When there are large changes in technologies, geographies, and patterns, in short periods of time, I think some skepticism may need to be applied to ALL data sources (think revisions to employment data for instance). That said, I continue to impressed by other indicator data, say the rig count. It hit 2000 a couple of times in this period, but I cannot vouch for Baker-Hughes character.