.comment-link {margin-left:.6em;}

Thesis & Antithesis

A critical perspective on energy, international politics & current affairs

My Photo
Location: Washington, D.C.

greekdefaultwatch@gmail.com Natural gas consultant by day, blogger on the Greek economy by night. Trained as an economist and political scientist. I believe in common sense and in data, and my aim is to offer insight written in language that is clear and convincing.

11 July 2006

Coping with high oil prices

The most surprising feature of the current oil crisis is that it does not really feel like a crisis. Oil and gas prices may be high and many people are struggling to cope with rising energy bills, but at a macro level, the world’s largest economies have grown consistently in the past two years. Hardly is our fear realized—that high energy costs will force an economic downturn, much less a recession. What explains this disconnect between expectation and reality?

To examine this question, take three mechanisms through which oil prices affect economic performance (there are more, but let’s focus on three): a reduction in income caused by the need to spend more money on energy leads to reduced demand for goods and services and this, in turn, forces an economic slowdown; an increase in inflation generated by higher prices that firms charge to cover energy costs leads to a reduction in real income; and worsening performance by firms, reflecting mainly higher costs and/or reduced overall demand by shrinking real income.

Begin with the first mechanism: US households spend more money on energy today—that much is obvious. But as Figure 1 shows, the amount spent on energy as a percentage of personal consumption is not very high, certainly not as high as it was in the late 1970s or early 1980s, when over 9% of personal consumption went to fuel oil, coal, electricity and gas. In 2005, energy expenditures as a share of consumption were 5.8%, up a full percentage point since 2003, but still below peak levels. Granted the data is not unambiguous—for example, current consumption is linked to high debt levels—but there is a clear sign that today’s high energy prices are not putting a strain on the economy that is comparable to that felt in the late 1970s.

What about inflation? Over the past fifteen years (1990-2005), there has been an observable decline in inflation, driven by a variety of factors unrelated to oil. As Figure 2 illustrates, the link between higher oil prices and inflation is all but clear (here is shown the Consumer Price Index excluding energy prices to gauge the effect of energy prices on other goods): in 1999-2001 there seem to be a rise in oil prices that is followed by an increase in inflation; after 2002, however, oil prices go higher, as inflation goes down; and by 2003, oil prices skyrocket, with only a minimal effect on inflation. Here, again, there are various exogenous factors to consider—mainly better macroeconomic management and the influx of goods from China, which have kept prices low. But this does not negate the underlying fact—that higher oil prices have not generated inflation, at least not to the level expected (and feared) by observers.

This reality produces the following question: could it be that firms are taking the hit? It is possible that firms would try to absorb energy costs in order to maintain demand for their goods. If this were true, we would expect firms that need a lot of energy to suffer more than firms that need less energy for their outputs. There is some evidence for this hypothesis, though the verdict is ultimately mixed: as Figure 3 shows, utilities and transportation—two energy intensive industries, had mixed results in 2004 with the former turning a profit while the latter suffering losses. From the rest (excluding Agriculture and Mining), there appears some trend, albeit weak, linking higher energy intensity and lower profits. At the same time, the numbers involved (energy costs at around 5-10% of total intermediate costs, and generally high profits) for most industries suggest that we cannot rely on this explanation—that firms are taking the hit—for understanding why oil prices are not having a large effect on the economy.

Perhaps the clearest view on this question comes from a more basic statistic: how much do firms spend on energy? Figure 4 shows gross output in the United States for the years 2000-2004 (gross output is labor and capital expenses, which make up GDP, as well as firm expenditures on energy, materials and services). The left axis plots gross output while the right axis shows total energy costs. What is impressive is that energy costs make up such a small portion of total costs (or total output). Even with high prices in 2004, energy costs make up about $450bn or 4.5% of total input costs. More than anything else, this should underscore why large changes in energy expenditures are not placing as high a strain on the economy, even though the fact that prices have risen more orderly than in the past may help explain why the adjustment has been less painful.


Granted, economic performance is but one aspect of the current energy crisis; it may not even be the most important. Granted too, that these numbers rest on a macro-level analysis and may conceal many problems, not least that of that families trying to pay their energy bills. Granted also that there are many international dimensions (even imbalances) to consider that may be salvaging economies from recession. But there is still some truth in here—that economies can grow in spite of high energy prices should make us rethink energy security and the calamities we tend to associate with rising oil costs. It may also give us some reassurance about our ability to make the transition from hydrocarbons to other energy sources as painless as possible. And that is good news for the long term.

Figure 1 data come from the Economic Report of the President (February 2006); Figure 2 data come from FRED—the Federal Reserve Economic Data; Figure 3 and 4 data come from the US Department of Commerce, Bureau of Economic Analysis, while the WTI spot prices for figure 3 are from the Energy Information Administration. All numbers / years are latest available.

Labels: ,


Blogger Nikos Tsafos said...

A shorter piece that consolidates the views expressed in three previous posts.

10:14 PM  
Anonymous Daniel said...

What about the roles of energy conservation measures and energy efficiency technology as factors in keeping the energy costs of firms low (i.e., for your second and third points)? Would these effects be large enough to appear at this macro scale?

11:59 AM  
Anonymous Anonymous said...

in Figure 1, what I find startling is that current personal expenditures on oil--even including those high debt levels--is just short of the low-point in the mid 1970s. If credit-financed consumption drops as consumers realize their energy expenditures have risen permanently or for the medium term, they probably will reach those earlier levels.

Also, do you have the percentage of energy costs of total input costs for earlier periods? That would make Figure 4 more convincing.

12:30 PM  
Anonymous BabyNutcase said...

What about contracts and hedge funds that are keeping firms from having to pay the spot prices? How long will the price increases in energy take to negate this effect as contracts expire and have to be re-negotiated?

1:22 PM  
Blogger Tony said...

I don't think energy costs can be broken out this easily. What about the energy costs embedded in the labor, capital, materials and services costs, to companies? When those are taken into account, what is the percentage of a companies costs directly or indirectly attributable to energy?

However, the conclusion misses the point that it is not so much rising energy costs that are the key factor for the future, but the decreasing energy supply, relative to demand. It's not clear that that has yet happened to any great degree. When it does, not only are costs likely to sky-rocket (until destruction of demand, i.e. recession) but yearly energy increases will stop.

Can we make any kind of adjustment without ample energy?


8:33 PM  
Anonymous Lawrence said...

The official US CPI figures that are used are essentially fraudulent.

You can read all about the manipulation of the CPI and other critical indicators at John Williams' web pages: http://www.shadowstats.com/cgi-bin/sgs


9:44 PM  
Blogger Nikos Tsafos said...

Daniel, energy efficiency and conservation are hard to capture at this macro scale. A proxy for energy intensity is easy to create but would not illuminate too much, particularly as the data do not go back many years (most of my raw data only go to around 1997-1998).

About hedging, I am not sure the effect would be that large. Most firms will only hedge maybe a year ahead, so they will still be buying contracts at higher prices; also, the market is in contango with future prices higher than spot prices meaning that futures contracts provide limited comfort. I am also thinking that the macro effects would probably largely cancel out (I could be wrong on this) since hedges are swap contracts where one party “wins” and one party “loses.”

It is true, as Tony writes, that there are other hidden energy costs which are not easy to measure, but I don’t think that diminishes the main idea—after all, most data contain linkages that are hard to measure. For example, it may be that energy prices affect costs of materials, but material costs also affect energy prices. It is not obvious to me how to disaggregate these effects.

I am also not sure that “the conclusion misses the point that it is not so much rising energy costs that are the key factor for the future, but the decreasing energy supply, relative to demand.” Prices will ultimately reflect an underlying supply and demand balance. My point here is that we should probability rethink our assumptions about what kind of oil prices we can live with.

As for fraudulent CPI data, I can only say one thing: I clicked on the link and it seems that John Williams takes issue with the level of inflation (which he estimates as higher) not its trend (which follows official government data). Using the alternative CPI would do nothing to change the correlation between it and oil prices since all changes in direction and magnitude follow the same pattern as official government data.

11:38 AM  

Post a Comment

Links to this post:

Create a Link

<< Home