In Understanding Business Fluctuations Not all GDP is Equal

by on December 28, 2017 at 7:25 am in Economics | Permalink

In standard macroeconomic models, GDP is GDP and an industry is important only to the extent that it produces GDP. In other words, there are big industries and small industries but there are no special industries. The standard view implies that the structure of production can be ignored. It doesn’t matter, for example, whether an industry sells to final consumers or to other businesses. It doesn’t matter whether an industry sells an easy-to-substitute product or a hard-to-substitute product. And it doesn’t matter whether an industry is a weak-tie link or a strong-tie link. Thus, in the standard view, the oil industry is no more important for understanding economic fluctuations than say the retail sales industry, since they are about the same size.

The standard view isn’t arbitrary, Hulten proved that it was true under certain assumptions. Hulten’s theorem appeared to offer a very useful simplification and so it diverted the attention of economists away from trying to model the structure of production.

But I’ve always been skeptical. One reason is that rapid increases in the price of oil have preceded almost all U.S. recessions (see Hamilton’s papers) and such increases appear to be much more important than the size of the oil sector would allow. More generally, although the canonical real business cycle model emphasizes aggregate technology shocks, I’ve always looked favorably on variations with sectoral shocks–this is one reason why Modern Principles is one of the few principles of economics textbooks to devote significant attention to real shocks, including oil shocks, and the mechanisms that can transmit and amplify these shocks from one sector to the wider economy.

The sectoral approach gets new support in a recent paper, The Macroeconomic Impact of Microeconomic Shocks: Beyond Hulten’s Theorem (non-gated) by Baqaee and Farhi. The authors show that Hulten’s theorem offers a true simplification only under very restrictive conditions. Relax those conditions and what seem like second-order mechanisms can have first-order effects. The structure of production–how industries are linked to one another, elasticities of substitution, reallocation speeds and so forth–matters in theory.

In a calibration Baqaee and Farhi consider “the response of GDP to shocks to specific industries” and:

…It turns out that for a large negative shock, the “oil and gas” industry produces the largest negative response in GDP – this despite the fact that the oil and gas industry is not the largest industry in the economy.
Hulten’s simplification came at too high a price. Further progress in understanding business fluctuations will come from developing microeconomic foundations to macroeconomics–not simply by subjecting a utility-maximizing representative agent to an optimization problem but by investigating the real structure of production in all of its complexity.

1 Dots December 28, 2017 at 7:38 am

Is there another discipline that has done better work than económics for our understanding of the economy? Economic sociólogy? Economic history? Industrial history?

Should i look around in French for this literature? histories of Embraer in Portuguese? Should i learn Germán, or some E. Asían languages?


2 Matthew J PhD December 28, 2017 at 7:50 am

The problem with this approach isn’t so much the calibration method but the fact that the paper he cites used a discontinuity regression approach.

Harvard PhD


3 Warren Platts December 28, 2017 at 12:58 pm

@Dots: Yes, economic history, and mercantilist theory in general have always argued that all sectors are not created equal. Don’t know about the French literature, but google Friedrich List. Also, Alexander Hamilton’s “Report on Manufactures” argued that the manufacturing sector is especially important for a country’s economy.


4 ohwilleke December 28, 2017 at 9:22 pm

I would argue that the subfield of economics known as “descriptive economics” is underdeveloped relative to its usefulness.

Lawyers and business school professors and business journalists at least have tools like a solid understanding of how the underlying business transactions work and an understanding of how complex economic enterprises are actually conducted (along with the conventional wisdom in those industries about how the economy works) which is a prerequisite to doing good work understanding the structure of the economy. This is something that economists tend to be embarrassingly uninformed about, as too many of them deal with the economy only at an abstract and theoretically driven mathematical model level.

Within economics, the folks who do econometric models, especially in microeconomic disciplines, probably have the firmest understanding of how the real economy, as opposed to the toy models that most economists prefer, actually works.

Economists are like biologists who understand DNA but taxonomy.


5 Merijn Knibbe December 28, 2017 at 8:01 am

Didn’t we call this, in the merry days before DSGE models (which treat GDP, wronglym as a top down governed homogenous lump of ‘something’ instead as of the ever changing bottom up hodgepodge which it really is) and using input output analyses, ‘backward and forward linkages’?


6 Euripides December 28, 2017 at 8:02 am

Nice summary, Alex. Indeed seems like an important issue. One barrier will be, how many sectors do you include exactly to make the model tractable in a macro framework….Well, let’s let the GMU and Minnesota grad students write some dissertations on this.


7 Warren Platts December 28, 2017 at 12:59 pm

God noooooo……


8 rayward December 28, 2017 at 8:06 am

Not to disagree with Tabarrok, but the biggest oil shock, the oil crisis (embargo) in 1973, affected behavior as much as or more than the macro-economy (the macro-economy was definitely affected, but as the result of behavior rather than the embargo itself). I appreciate that behavior is difficult if not impossible to separate from the direct economic impact of the embargo, but having lived through the crisis my observation is that it significantly altered behavior. One should also remember that the price of oil before the embargo was only $20 a barrel as opposed to the “depressed” price of a little under $60 per barrel today, which is higher than the price reached at the height of the embargo (these are inflation-adjusted figures). “Let us do what is evil so that good may come” has special significance for the oil crisis, as it greatly increased efforts to find alternative sources of energy and greatly increased the energy efficiency of industry, housing, and transportation, which in combination greatly increased macro-economic output.


9 Benny Lava December 28, 2017 at 8:21 am

I remember in 2008 arguing with twats like Megan McGargle that the big spike in oil prices was going to crater the American economy. I am no genius or oracle but my argument at the time was if you raised taxes on everyone by a few thousand dollars a year wouldn’t you expect a contraction?

Anyways good to see Alex post something that agrees with empirical evidence. Useful post.


10 Jeff R December 28, 2017 at 10:54 am

Oil prices brought down Bear Sterns and Lehman Brothers?


11 Warren Platts December 28, 2017 at 1:19 pm

People who couldn’t afford to drive to work stopped making their house payments. Also, increase in fuel entailed a decrease in foreign investment in US debt instruments as foreigners reallocated investment monies into oil purchases. So the trade deficit was also involved.


12 Floccina December 28, 2017 at 11:57 am

But those in the oil industry get that money.


13 Scott Gustafson December 28, 2017 at 2:07 pm

Most of what you pay for a gallon of gasoline goes to a government somewhere.


14 rayward December 28, 2017 at 8:29 am

To appreciate how much has changed, before the 1973 oil crisis gas sold for less than 30 cents per gallon, which was a good thing because cars (my car anyway) got only 6-8 miles per gallon. I would buy one or two dollars of gas at a time, the gas pumped by the guy who worked at the gas station (who would check the oil, the fluid in the radiator, and the air pressure in the tires) not by the customer because there was no such thing as self-service. A stroke of luck for me, a year or two before the oil crisis I sold my gas guzzler (a Pontiac GTO) and bought a Datsun (n/k/a Nissan), which was more like a go-cart with doors than a car. American cars back then were terrible, which is why people traded so frequently. I paid less than $2,000 for my brand-new Datsun. But that price didn’t include an air conditioner because Datsun didn’t offer cars with factory-installed air conditioners.


15 rayward December 28, 2017 at 8:39 am

Busch Beer sold for 89 cents a six back. Plus the deposit (it was sold in returnable bottles). For someone in my financial position at the time, the returnable bottles were a form of savings: when short of gas money, I would return the bottles for a refund of the deposit so I could buy a dollar or two of gas. I was a very good saver back then. I made Tabarrok and Cowen proud.


16 Abe Simpson December 28, 2017 at 12:02 pm

…I wore an onion on my belt, which was the style of the time.


17 David Pinto December 28, 2017 at 8:56 am

There was a time, the late 1990s into the early aughties that I thought the Fed should leave economic policy to the oil industry. It seemed that when the economy heated, oil prices rose and served as a break. When the economy pulled back, oil prices dropped and the economy could move forward again. Maybe I was right and oil is a better modulator of the economy than the Fed.


18 Dick the Butcher December 28, 2017 at 11:54 am


I thought the purpose of the 1913 Federal Reserve Act was to mitigate or abolish the business cycle.

Alternatively, it has been said that copper is the commodity with a PhD in economics.


19 Jeffrey Deutsch December 28, 2017 at 1:52 pm

Are you saying we might even put the dollar on the oil standard?


20 ohwilleke December 28, 2017 at 9:37 pm

What I find appalling is how many economists think that the only thing with knowing about the macro-economy is how the money supply, and interest rates as influenced by federal reserve policy, influences the business cycle. This is one small story in the big picture, but it isn’t the more important or the only thing worth knowing about the macroeconomy.

For example, there is an important story to about told about how, during the Great Recession, the collapse of a couple of key state housing bubbles driven by bad state foreclosure laws upset the entire mortgage backed securities market which reached a tipping point that led to a crisis in the entire financial system that in turn hurt unrelated industries like manufacturing in places remote from the housing bubbles that started it all. Bad foreclosure laws in California through a chain reaction effect, damages a swath of small factory owners in Tennessee.

These stories, which are true ones, can only be understood in the context of a world view nuanced enough to understand the more important strands that link different sectors of the economy and an awareness of where the weak links in the chain are located. For example, I remembers sitting in my securities law class in 1993 thinking to myself – the entire bond market depends almost entirely upon a handful of lightly regulated judgment proof bond rating agencies with very little skin in the game accurately assigning credit ratings to fixed income securities, and that if that link failed to perform (as it seemed very vulnerable to doing) that these firms’ bad judgment could take down the whole financial sector. And, fifteen years later, in 2008, this is exactly what happened.

Of course, knowing this as a young law student and as a newly minted medium sized law firm lawyer far from the nation’s financial centers, provided very little opportunity to do anything about this knowledge for good or ill, for either myself or my clients.


21 Dave M December 28, 2017 at 9:03 am

Nice summary, Alex. I’m not an economist, but it would seem that not only structure but context should matter a great deal, especially for a sector like oil an gas. Oil and gas is traded on markets, used in transportation and energy, and utilized in the production of raw materials. Add in the politics in the US that surround its production – ANWAR, Keystone Pipeline, competition with petro-states – and it would appear rational that oil and gas GDP isn’t on the same footing as other sectors.

Of course, you could probably point to every other sector and find just as much complexity!


22 ʕ•ᴥ•ʔ December 28, 2017 at 9:16 am

Oil is more than a sector, though. It is energy, and thus fundamental to all substitution of labor with machine.


23 Brian Donohue December 28, 2017 at 12:02 pm

Seems pretty obvious to me, conventional wisdom even. Also, energy is less fundamental than it was 40 years ago.


24 ʕ•ᴥ•ʔ December 28, 2017 at 12:14 pm

Yes, reduced energy intensity has been a great benefit.


25 Floccina December 28, 2017 at 12:03 pm

Only 37% of energy used in the USA comes petroleum.


26 ʕ•ᴥ•ʔ December 28, 2017 at 12:05 pm



27 Floccina December 28, 2017 at 12:06 pm

It looks like personal transportation is about 45 percent of total U.S. oil consumption.


28 ohwilleke December 28, 2017 at 9:44 pm

“It looks like personal transportation is about 45 percent of total U.S. oil consumption.”

This sounds low to me. The only places that oil is used in any significant way to produce electricity are Hawaii and Alaska. The only place that oil is used by households for purposes other than personal transportation is in the Northeast where it is used for heating oil (and the oil they use is a low quality heavy oil component relative to the component of oil used for transportation fuel which has mostly shorter carbon chains than heating oil). Plastics, fertilizers and airplane fuel are pretty tiny shares of the total oil pie. And, industrial oil consumption (along with industrial coal consumption) has declined steadily for decades. Motor vehicles (mostly cars and trucks) are the predominant users of oil today.

The only way I can see that you get to 45% on personal transportation, is if most of the rest of “business transportation” involving functionally identical cars and trucks and construction vehicles running on gasoline and diesel as well.


29 ʕ•ᴥ•ʔ December 28, 2017 at 12:13 pm

Ok, a more serious answer, and not to go all peak oily on you ..

Forbes says that a barrel of oil contains about the same amount of energy as a human would expend in calories in 11 years of manual labor.

Per your link, that 37% represents 18,690,000 barrels per day.

That is a lot of human equivalent labor, 205,590,000 years? compressed into every day.

Shortage or price increase has to hurt.


30 Anon December 28, 2017 at 10:07 pm

Come on, equal energy is not the same as equal impact


31 ʕ•ᴥ•ʔ December 28, 2017 at 11:15 pm

Just trying to thumbnail how much beyond human/animal bodily energy a modern technological civilization requires.

You wouldn’t have Costco without *millions* of times the energy of an animal economy.

It would be nice if we could do even better, with cheap fusion, solar housepaint, or whatever.

32 Steven Kopits December 28, 2017 at 3:26 pm

Yes, at least with respect to transportation. Not necessarily for manufacturing, much of which can be run on gas or electricity.


33 Bill December 28, 2017 at 10:20 am

Better explained with behavioral macroeconomics where there is a salient shock viewed as a loss which causes rapid readjustment in spending behavior to maintain wealth, income, or social position.

Look up some articles on behavioral macroeconomics and supply or demand shocks.


34 Anonymous December 28, 2017 at 1:21 pm

I’m skeptical that behavioral macroeconomics is required. A supply shock will make marginal industries/activities that use that commodity as an input unprofitable, and if those industries are capital intensive, the that capital will suddenly become worthless. The total capital intensity of the economy will decrease, decreasing it’s real output potential.


35 Bill December 28, 2017 at 8:04 pm

If you believe that everyone is rational, then someone rational would have hedged against a risk, or would not overreact to it. Moreover, remember, we are not dealing with financial gurus here and their decision making under uncertainty and ambiguity, but stupid folks, of which there are many, who overreact to transitory, salient events.

You might be interested in reading this about behavioral macroeconomics and shocks:


36 Jeffrey Deutsch December 28, 2017 at 1:56 pm

Behavioral macroeconomics means we’d need to focus on what John Q. Public believes to be the “commanding heights” of the economy.


37 Tom T. December 28, 2017 at 10:29 am

The problem with Alex’s postulate is all of times when a sharp run-up in oil prices did not result in recession (the mid-80s, the mid-90s, most of the 00s*). The evidence seems to fit better if one runs the causation arrow the other way — recessions cause the price of oil to drop.

* Yes, this run-up eventually terminated in the 2008 crash, but if the theory has to be extended to “a sharp rise in oil prices means there will be a recession sometime in the next seven years,” then its predictive power is effectively zero.


38 Steven Kopits December 28, 2017 at 3:23 pm

Oil prices were low in the 1980s after OPEC capitulated in the 1983-1985 period, and they remained low until the end of the Great Moderation around 200-2005.

Jim Hamilton has an ‘inoculation’ theory (my words, not his), that an oil price already seen by an economy will not cause a subsequent recession. I think there is some truth to that, but I might be less categorical in my view.


39 Steven Kopits December 28, 2017 at 3:27 pm



40 Ol' George December 28, 2017 at 10:41 am

This post is a rather long-winded way of saying “I don’t understand how sticky prices/wages work”.

Is this guy supposed to be part of “the best and the brightest” ?


41 Anonymous December 28, 2017 at 1:27 pm

Or he’s entertaining alternative explanations?


42 Bob is Speaking to Idiots. December 28, 2017 at 2:51 pm

No, he is not — Tabarrok is barely an economist. I wonder who read him Baqaae and Farhi’s paper.


43 Steven Kopits December 28, 2017 at 3:30 pm

No. The key is the volume, not the price, of oil.

If the economy is not able to obtain the volume of oil necessary to function, then production and consumption must be rebalanced, ie, fewer cars, fewer vacations, smaller houses, to be replaced with potentially more services and intangible goods. That realignment of the economy can cause a recession. The recession may be prolonged by sticky wages and prices, but the underlying cause is still a real supply shock.


44 Ol' George December 29, 2017 at 10:56 am

Thank you Captain Obvious


45 TallDave December 28, 2017 at 10:45 am

Interesting result, thanks for sharing. Obviously an oil supply shock in the age of fracking is a different beast (at least in the US), would be interesting to see how much different at similar ratios of supply.

I was going to download the paper, but the NBER does not agree Illinois is a developing or transition economy (clearly they need to visit!). Intriguing summary, anyway.


46 Warren Platts December 28, 2017 at 1:25 pm

There’s a link to the free version in the OP next to the NBER link.


47 Al December 28, 2017 at 11:00 am

Fracking, which is despised by the left, is responsible for the prosperity we see today.

I would really enjoy a simulation by an economist as to the state of the world if we had banned fracking as all good liberals wanted. Sadly, all modelers are employed in climate studies. Strange.


48 ʕ•ᴥ•ʔ December 28, 2017 at 11:11 am

You an idiot. Anyone else on this page understands that fracking has risks and rewards, any that it is not an abstract or ideologically pure morality play.


49 Warren Platts December 28, 2017 at 1:42 pm

Correlation does not equal causation. Note that your study did not control for household income. Probably people with a lot of money prefer not to live in the middle of a dense oil patch.


50 ʕ•ᴥ•ʔ December 28, 2017 at 2:31 pm

My point really is just the energy vs externality trade-off. I picked an example of that, and you might quibble with it, but I don’t think any reasonable person would say oil in general and tracking in particular is pure positive.

Hobbled only by “leftists.”


51 Steven Kopits December 28, 2017 at 3:17 pm

Fracking is almost entirely a positive. It’s about as benign a form of resource extraction as you’ll find.

To argue against fracking, you have to put a very high price on carbon.

52 ʕ•ᴥ•ʔ December 28, 2017 at 3:49 pm

My main concern is about process. Are we doing health and environmental studies scrupulously and honestly? Or are we (as I sometimes fear) dividing into “no study” and “stacked study” camps?

(I suspect that local water and air pollution is more likely to be an unmeasured risk than global CO2. We leave it that way, because we like cheap gas. Tough on the poor people, as Warren says.)

53 ʕ•ᴥ•ʔ December 28, 2017 at 3:50 pm
54 Steven Kopits December 28, 2017 at 4:27 pm

As I understand it, there are four risks associated with fracking:

1. Groundwater contamination from poor completions
If the piping associated with a well is poorly installed through the aquifer, generally the top 250 ft, then groundwater may become contaminated. (Fracking is usually conducted at depths around 4,000 ft.) This is a manageable risk, really related primarily to good manufacturing process. I’d add that any number of aquifers in Pennsylvania are naturally contaminated with hydrocarbons. To wit, oil was discovered at a depth of 29.5 feet in Titusville, PA, back in 1859.

2. Leakage from retention ponds.
Fracking and produced water from wells may be stored in open retention ponds. If these ponds — which are ordinarily rubber lined — leak, then nearby areas, conceivably including groundwater, could be contaminated. This again is primarily a question of proper site management.

3. Water disposal
In general, produced water is disposed of using approved methodologies. It’s not inconceivable that some would be disposed of improperly. This is illegal, but illegal things can happen sometimes.

4. Earthquakes associated with produced water injection
A popular method of produced water disposal is re-injection into old depleted oil wells or disposal wells drilled specifically for that purpose. The data suggest that such practices can in fact induce small earthquakes, with Oklahoma a notable example. This can be mediated by appropriate management, as it largely has been in Oklahoma.

Fracking has been done for more than fifty years. It is well understood and not particularly dangerous, assuming proper procedures are followed. Such contamination as occurs will tend to be local, as opposed to an offshore spill, which can cover hundreds of square miles.

Really, compared to activities like mining or even conventional oil production, fracking puts much less strain on the environment.

55 ʕ•ᴥ•ʔ December 28, 2017 at 2:31 pm

“Fracking” not tracking.


56 Al December 28, 2017 at 6:57 pm

Thanks for proving my point about your camp.


57 ʕ•ᴥ•ʔ December 28, 2017 at 11:19 pm

Read Steven’s list again. He makes my point that there are recognized externalities.

He just makes the rather unsafe assumption that local risks are everywhere managed.

My links suggest that is what in question.


58 ChrisA December 29, 2017 at 3:28 am

If fracking is so problematic where are all the law suits? And don’t tell me big oil doesn’t get sued look at the Deepwater Horizon penalties. Fact is fracking is incredibly safe and opposition to it is the lefts version of anti-vaccers.

59 ʕ•ᴥ•ʔ December 29, 2017 at 9:22 am

Can you guys even read?

I talk above about the benefits of a mass energy economy. I talk here about the trade-off with externalities, which Steven helpfully expands.

The nutters are on the extremes, “horribly unsafe” or “harmless.”

But did you even Google for lawsuits?


60 Steven Kopits December 30, 2017 at 1:43 am

Yes, there are externalities, and this is not debated even within the oil business. If you don’t regulate drillers, they will be potentially careless about aquifers and indifferent about waste disposal. But this is really no different than, say, a manufacturing plant or a meat processing facility. You need adequate, but not excessive, regulation and enforcement.

I think enforcement may be weaker in areas with a tradition of production and a large number of remote, rural locations. For example, Texas oil culture really evolved on the standards of a century ago, and these standards may tend to persist simply because that’s the way things have always been done.

In any event, fracking can be done safely, but no one should assume that drillers are the Mary Poppins of American commerce. Drilling activities do need competent and judicious regulation and enforcement.

61 ʕ•ᴥ•ʔ December 29, 2017 at 9:28 am

This is mostly about people who collapse odds to 0 or 50 or 100%, who can’t handle “odd” odds, let alone uncertain odds, as being real.


62 Matthew Young December 28, 2017 at 11:04 am

If you want structure, you have to obserserve congestion. This is what traders do with their charts; technical analysis.


63 shrikanthk December 28, 2017 at 11:44 am

A total aside.

But talking of oil shocks, is it fair to say that the inflations induced by oil cartels in the 70s are a refutation of the monetarist orthodoxy that inflation is created only at the mint when the government prints too much money and as a result too much cash chases too few goods?

As per monetarist theory, oil shocks should have merely meant reduced demand for other goods / services, and hence no aggregate increase in price level. Correct? But then we did have sustained high inflation rates in the 70s. I am not sure how much of the inflation can be attributed to Oil shocks vs loose monetary policy.

Thoughts anyone?


64 ʕ•ᴥ•ʔ December 28, 2017 at 11:54 am

I think this was a big miscommunication between academics, through the economic press, and to lay people.

Academics may know that a price shock can show up in CPI or PPI, but it will be reported on the nightly news as “inflation.”

Should price indexes soar again, this will probably repeat.


65 Brian Donohue December 28, 2017 at 12:06 pm

It was mostly monetary policy.


66 Anonymous December 28, 2017 at 1:38 pm

I would say that they decrease the real output potential of the economy, so monetary policy could still prevent inflation by tightening, but that may cause more dislocation in other sectors than allowing higher inflation.


67 Scott Gustafson December 28, 2017 at 2:14 pm

The reactions of the Fed was to continue increasing the money supply even though real output fell. That’s inflationary.


68 Benny Lava December 28, 2017 at 2:37 pm

Scott Sumner has argued that the fed created the inflation of the 70s. As an example there was a big spike in oil prices in 07-08 and no corresponding inflation.


69 Jeffrey Deutsch December 29, 2017 at 10:53 am

As per monetarist theory, oil shocks should have merely meant reduced demand for other goods / services, and hence no aggregate increase in price level.

That’s what I assumed while studying under Tyler Cowen & Company (back in the ’90s).

I think now that this implicitly assumes most if not all relevant externalities are pecuniary.

But what if quantity changes remain quantity changes through the system, or what if price changes translate into quantity changes?

Market power, with downward sloping demand curves and less than fully elastic — or even inelastic — demand and supply. Sticky prices. Sticky wages.

Then increased oil and gas prices (or decreased supply) to U.S. businesses don’t just mean some prices go up relative to others and quantities demanded shift. It may mean that actually less gets produced…which may even also cause stagflation.


70 B.B. December 28, 2017 at 12:11 pm

Why does the size of the oil industry matter? The US gets the majority of its oil from abroad. Large spikes in oil prices preceded US recessions. True enough, but that has nothing to do with the US oil industry.

Countries that have zero domestic production of crude oil or oil refineries can get hit very hard by an oil price shock.

The issue is consumption of oil, not production of oil, and how essential oil is to other industries and consumers.


71 Warren Platts December 28, 2017 at 1:52 pm

US imports about 10 MMbbl/day (mostly crude), but it exports about 6 MMbbl/day (mostly refined petroleum products); meanwhile, we produce about 10 MMbbl/day. Thus imports make up less than 30% of total US domestic consumption. Production matters. The more oil we produce, the bigger cushion we have against international oil shocks.


72 spencer December 28, 2017 at 2:04 pm

That was once true, but now US net oil imports are a marginal source of oil.


73 Steven Kopits December 28, 2017 at 3:14 pm

Actually, US net crude and product imports over the last four weeks averaged 3.0 mbpd, on consumption of 20.0 mbpd, representing an import dependence of 15%. We’ve come a long way from the days when we imported 60% of the oil we consumed.

If a country produces its own oil, then an oil price spike will help the oil sector even as it hurts the manufacturing sector. Ergo, during the period of high oil prices, the middle of the country from Houston to North Dakota boomed even as the coasts struggled. Without oil production, everyone would have struggled–which was pretty much the European experience.


74 Tom Barson December 28, 2017 at 12:44 pm

I don’t have a problem with decomposing macro models into sectors, or distinguishing between intermediate and final goods – go right ahead if the results are interesting. What I don’t understand is the last paragraph: why do we need to claim the term ‘microfoundations’ for this kind of decomposition? And in the case of oil prices – isn’t this confusing a ‘foundation’ with a shock?

Even if we decide we want to endogenize the energy sector or energy prices in some way, what does this have to do with whether a model uses a “utility-maximizing representative agent” on the consumer side? (Or did you mean a representative firm? But you would still have to say something about firm behavior if you had an array of firms or sectors.) There are multitudinous criticisms of the representative agent approach, but I don’t see what it’s got to do with accounting for intermediate inputs on the production side.

But I agree the Baqaee and Farhi research cited is interesting. Maybe for some kinds of questions it’s a needless complication to decompose the production side. but in other cases it could be very important.


75 KevinH December 28, 2017 at 12:53 pm

A good example of where studying network theory and fluid dynamics is probably a smart way for young economists to move.

e.g If the main article is true, then the max-flow min-cut theorm probably can tell you a lot about barriers to economic growth


76 Philo December 28, 2017 at 1:23 pm

“Thus, in the standard view, the oil industry is no more important for understanding economic fluctuations than say the retail sales industry, since they are about the same size.” But who holds this “standard view”?

It is trivially true–true by definition–true as a matter of accounting–that a decline in production from one period to the next of P percent in one sector of the economy contributes equally negatively to GDP with a decline of P percent in any other sector of the same size. Indeed, if we look at the changes in absolute value of production in different sectors rather than at percentage changes, we may drop the same-size limitation: a decline of X dollars in one sector contributes equally negatively to GDP with a decline of X dollars in any other sector, regardless of their relative sizes.

However this point about accounting tells us nothing about causation, which (I presume) we would have to grasp in order to “understand” fluctuations in GDP. An X-dollar decline in one sector, such as oil, may cause subsequent declines in other sectors, while an X-dollar decline in some other sector, such as retail, would have no such negative effect. But who ever denied this?


77 Freidrich December 28, 2017 at 2:02 pm

You could come up with a purely monetarist explanation. High or rising oil prices make the Fed more skittish about inflation -> tight money -> recession. Doubt it always applies but I do think it goes some of the way to explaining why they dropped the ball so hard in 2008


78 Steven Kopits December 28, 2017 at 2:08 pm

We refer to a certain set of goods as ‘enabling commodities’, as they enable other forms of economic activity.

Oil is one, but so is water and air, for that matter. If there were no air — a commodity which we don’t even measure as having value — GDP would fall to zero. Similarly, droughts and resulting famines are the stuff of historical disaster.

Oil is important as the de facto monopoly fuel for transportation. Transportation is necessary to move goods and people from production to consumption. A shortage of oil reduces our capability to produce and consume, with the regulator being the pace of efficiency gains. Obviously, if efficiency can be increased easily, then oil consumption can be readily reduced without material impact on economic activity. If not, then economic activity — or certainly its growth — will be restrained. This could result in ‘stagflation’ or ‘secular stagnation’ as the case may be.

I have a long presentation on this I made at Columbia University back in 2014. Still a cult classic.


79 spencer December 28, 2017 at 2:10 pm

The price elasticity of demand for oil is among the lowest in the economy.
So an oil price spike causes demand for other thing to be impacted much more than a similar price spike for other goods or services would impact other goods or services.


80 Paul December 28, 2017 at 2:16 pm

There’s a confusion here. “In macroeconomics”? Which part of macroeconomics?

In macroeconomics “producing GDP” can mean:

1. Contributing to long-run growth, as in, are we richer as a country.
2. Contributing to business cycles.

Those are different questions.


81 Scott Gustafson December 28, 2017 at 2:17 pm

Prior to 1973 the US economy spent 1-2% of GDP on crude oil. After two price shocks, in about 1981 it was a bit over 5% of GDP.


82 Scott Gustafson December 28, 2017 at 2:20 pm

Today we’re back at about 1-2% again.


83 Steven Kopits December 28, 2017 at 3:35 pm

Oil consumption is currently 2.3% of US GDP and about 2.4% of world GDP.


84 Boonton December 28, 2017 at 3:08 pm

One idea to consider, perhaps oil & gas are the most rational of sectors in the economy. A spike in oil/gas prices then might be the indication that even the supposedly sane analysts have drunk the bubble kool-aid and are buying into absurd levels of demand in the near future. This demand will never actually materialize, though, so perhaps oil/gas is the last to spike and first to crash signaling the start of the recession.

Why should this particular sector have the honor of being so ‘grounded in reality’? Possibly because it’s so much ‘on the ground’….people drive vehicles and people have to drill for oil/gas, in the very short run this doesn’t change all that much but since there’s little flexibility in short term supply even minute deviations from expectations cause wide price swings.


85 Jon December 28, 2017 at 3:57 pm

The energy sector represents about 46% of capital investment–about half of that is in oil and gas. That means the price of oil and gas is highly coupled to the natural rate of interest and does much more to swing a given fed target in to being erroneously easy or tight than variation in the retail sector which is capital light.


86 ohwilleke December 28, 2017 at 9:50 pm

Fascinating. I learned something new today. This fact illuminates a lot of economic and political-economy mysteries.


87 Warren Platts December 28, 2017 at 4:24 pm

Hmmm… So if it is the case that some sectors are more important than others, then it could possibly be the case that the goods manufacturing sector is more important than the burger-flipping service sector. In which case, the unilateral free trade that exchanged the former for the latter may not have been the brightest idea ever.


88 Philo December 28, 2017 at 11:47 pm

. . . more important *for understanding economic fluctuations*–not “more important” in some absolute sense.


89 Benjamin Cole December 29, 2017 at 3:01 am

The Fed seeing higher oil prices tightens the money supply leading to recession..


90 ChrisA December 29, 2017 at 3:24 am

Right. This is standard text book, High oil prices are a sign of a strong economy because oil consumption is rising faster than anticipated by investors in oil, so it is a good sign that monetary policy is loose.


91 BJ dubbS December 31, 2017 at 11:24 am

You can always tell it’s a Tabarrok post because it doesn’t make allusive, shamanic statements with no verifiable content.


92 Leigh Caldwell January 9, 2018 at 1:52 pm

Is it possible to identify in Hamilton’s data times when the oil price rise is the *cause* of the slowdown, and times when it is the canary in the coalmine? It seems plausible that if the output gap falls too low, or some generalised supply shock occurs, the oil price would jump first because demand is so price-inelastic (as is supply in the short run). But the oil rise wouldn’t itself be causing the slowdown, just signalling that it is coming. (I see that Hamilton mentions this possibility but thinks it’s incorrect, “at least for the early postwar period”)

Not that Baqaee and Farhi’s work isn’t also plausible – it makes a lot of sense that industries that are harder to substitute from can have more (negative) impact on the economy.


93 Jacques René Gigère January 18, 2018 at 4:53 pm

In an input-output matrix, oil appear in all columns. Anything that impact it will have a disprportionate effect ewverywhere.
But this technique, essentially invented in the US for finding what to bomb first and then used by Europeans to identify what to rebuild ( the same list in reverse), while common tool amongst European and Canadian economists is now vitually unknown in the US (it reeks of “central planning”).
A few years ago, aprominent US economist was keynote speaker at a large gathering of Québec economists. He began his talk thus:” If I was a Québécois or Canadian economist, I’d use an input-output model and be done in 20 minutes. But as I am american, I’ll take 2 hours.”


Leave a Comment

Previous post:

Next post: