« Obama vs. Huckabee vs. Bloomberg? | Main | Huckabee Burns the Republican Formula »

January 02, 2008

Comments

Fabius Maximus

I disagree about the comparison of CERA's work vs. that on the The Oil Drum. CERA's data and analysis are consistently world-class, while that of TOD is very uneven. Some is excellent, some a joke (that's probably inherent in their structure).

As for forecasts, even dartboards often outperform experts. One trial tells us nothing. Also, most of the writers at TOD are perma-bulls on oil, inherently right as oil prices rise. If we have a recession which crashes oil prices, it wil be their turn to look foolish.

Note that oil prices have risen along with most commodities over the past 5 years, which should raise questions about the role of the "peak oil" factors TOD discusses vs. broader economic ones they tend to ignore(strong global economy, inflation).

gmoke

All this Fall, I have been asking folks who come to speak at Harvard's Shorenstein Center about the idea that Peak Oil may have already occurred, as theoildrum's production updates seem to show.

Nobody's answered intelligently yet.

If there are alternative world oil production figures to those on theoildrum I'd like to see them. I'd also like to see a newspaper article on oil prices that actually includes a world oil production graph within a historic context.


Fabius Maximus

The articles on Peak Oil typically use EIA/IEA data, which is industry standard. Unfortunately it is, like most global economic data, not very accurate (Simmons gives some great examples of this). Worse, I've seen no analysis indicating what the error bars might look like (such as you get with US data, like the jobs numbers). Who knows, the errors might be larger than the YoY changes.

The data on inventories is even worse. We have only data on primary inventories for the OECD nations, almost nothing on secondary inventories, nothing on tertiary (end-users'). So the actual consumption trends are apparent only after months -- or years. Who knows?

We are in the 5th year since oil prices began rising, when we might expect to see capital investments and usage changes (conservation and increased efficiency) effects in the data. Perhaps that accounts for the consumption plateau over the past few years.

To see what this might look like, consider the 15% drop in oil use 1979-83, growing back to the 1979 peak only after 14 years. During this period the global economy did just fine, despite the claims that a decrease in oil consumption means an inevitable depression (see presentations at the last ASPO conference for examples).

Fabius Maximus

gmoke - The Oil Drum is not a primary source. Try the EIA and IEA websites, which have a wealth of data. Also see BP's "Statistical Review of World Energy", an accepted and easy to use source.
http://www.bp.com/productlanding.do?categoryId=6848&contentId=7033471

John Robb

You don't have to agree with peak oil to get significant benefit out of the oil drum. The debate between alternative projections and the desire to attack underlying assumptions produces tremendous insight.

In contrast, CERA by design represents a single approach. So, if its projections are wrong, the entire product is suspect.

Fabius Maximus

Most of what people pay for with products like CERA is a reliable source of data and analysis. The forecasts are the cherry on the top, but nobody -- absolutely nobody -- has a decent record at forecasting such things (e.g., GDP, oil prices, stock market prices).

The Oil Drum is an excellent blog, with much interesting material (I have it on my blogroll). Such as Staniford's article using Hubbert Linearization to predict global post-peak decline rates (accelerating *very* slowly, after 40 years reaching 4%/yr).

This is mixed in with some nonesense, hence the need for caution -- esp for non-experts who might not be able to tell the difference.

To give just one example, note almost dogmatic assumption of most TOD articles that decreased oil consumption means falling GDP (working both ways). The 1979-93 global data shows that this is not inherently so, yet I've never seen this mentioned.

The comments to this entry are closed.