Interesting. So we have had two prior "false peaks." What occured in 1999 and 2002?
The first one is the Asian currency crisis, and the second is the tech crash.  Prices were not particularly high in either case (according to BP on an annual basis, WTI was $20 in 1997 and $14 in 1998.  In 2000 it reached $30).  It seems clear to me that both events are demand-side related.
Weird, we both look at it and come to 2 completly different conclusions at the same time!

One says supply and one says demand.

I'm no expert so I must assume you are correct.

Well, that's handy.  Calculated Risk  (posting at Angry Bear) has a nice summary of Fed transcripts from May 2000, right as things were about to go South.  Here's US GDP changes:

You can see that GDP slows down really a lot in late 2000, whereas the oil production in OilCEO's graph above doesn't slow down until the end of 2000.  So GDP changes preceded oil production changes.  However, in the current case, oil production growth started to slow down in late 2004, and GDP growth was good until Q4 2005.

Hence it appears that the causality is reversed in the two cases.

We'd have to look at Asian data to see the 1998 events, but I don't have it to hand (nor time to hunt now).

Stuart, I'v been futzing around with price all night trying to somehow add it to this graph. Unsatified at my attempts to somehow relate the two in a chart that shows promise for a correlation. Any thoughts?
In Excel, under the Chart Menu, choose "Chart Type".  Select the "Custom Types" tab, and then choose "Lines on 2 Axes".  

Also, if you've already got a chart you can add another series on a second axis somewhere under the Data menu but I forget where.

Thanks. Excel isn't the problem I'm having, it's conceptualizing the relationship between price and production.  They move percentage-wise in magnitudes that are incompatible, so you have to dampen the price magnitude by several factors.

I've tried indexing them to a "100" number but then how do you choose what month corresponds to "100" for both except for arbitrarily?

Maybe I'll post some of my best aborted attempts later so you can see what I mean.

Maybe there isn't anything relevent we can draw from the two? Maybe there is, I just can't get a graph to show it.

That's why you want two axis - it lets you have both variables on difference scales on the same graph.
Yeah, you're right, I'm gonna try that next.
Oil prices took a nose-dive from the previous year or 2.

From the $20+ range in 97, it lowered until bottoming out at near $9 in Feb of 1999.

In Early 2002 it bottomed out at $16 and was much higher years previous.

I think there was an overshoot by oil companies, at least in these 2 cycles, in which they brought to much oil to market.  And prices fell, so they lowered the amount produced to bring them back up.
This time looks different.

Re: "False Peaks"

We have seen this is in many places, such as Texas.  The production declines that count are the ones at or past 50% of Qt.  

Question:  what do these regions/countries have in common:

Texas; Lower 48; Total US; Russia; North Sea.

Answer:  Oil production from these regions has never exceeded the peak they saw in the vicinity of 50% of Qt.

The world, from conventional sources, is at 50% of Qt.

That is why I think it makes sense for Stuart to look at the detailed production numbers.  These small production declines--and small declines in imports (IMO)--are small pebbles falling down the mountain just ahead of a massive landslide.

Re:  Hubbert Linearizaton (HL)

The HL method allows us to estimate the area under the curve on a production versus time graph, which is total estimated recoverable reserves, or what Deffeyes calls Qt.   The assumption is that production peaks at about 50% of Qt, and historical records support this.  

IMO, for HL to really work we need a region with about two mbpd of production for about 20 years.

Unless one produces a field at a high enough rate to damage it, the production rate, within limits, tends not to have a major effect on cumulative production.  Therefore, a region, using a gross simplification, can be compared to a bottle full of water.  You can pour it out a a minimal rate all the way up to the maximum rate.  But the rate that you pour it out has no effect on the volume of water in the bottle.

The two largest discrete regions which have rolled over and gone downhill in the vicinity of 50% of Qt are the Lower 48 and Russia.  Russia had wildly inconsistent production data after the fall of the Soviet Union (in 1989), so Khebab and I just used Lower 48 production data through 1970 and Russian production data through 1984, to predict future production using HL

Using only production data through about 50% of Qt, the HL method was 99% accurate in predicting cumulative Lower 48 production through 2004 and the HL method was 95% accurate in predicting cumulative Russian production through 2004.  In other words, the water was poured out of the bottles at different rates, but the rates had no effect on the volume of water.

First, this suggests that Deffeyes' prediction that we have used 50% of total conventional oil reserves should be given a lot of credibility.

Second, this suggests that Russian oil production, and therefore its exports, and in fact the exports from the top four oil exporters, are going to be falling rapidly.  Thus, my focus on US imports year over year.  

Have we seen very sharp production declines?  Yes,.   Consider Cantarell (the second largest producing oil field in the world), where the remaining oil column is about 850', and it is falling at the rate of about 300' per year.

So given these numbers and thoughts.  Where and when do we see a World shortfall of production?  

If the USA does something with Iran whatever it is in the next year or at most 2 years, at least before the 2008 Presidental Elections.  Where will oil production worldwide go?  Down due to conflict, and stay down for years, and never recover. or what?

ps.

I hate being an information junkie with the DOD on my back.  Federal Law is a nasty bedfellow. Growl.

IMHO if Pakistan loses its current leader, They won't own the nukes anymore.  

"Where and when do we see a World shortfall of production?"

I think that we will see a drop off in oil exports before we see a large overall decline in world oil production.

I think that are seeing a drop off in net oil exports right now. I know that the data are noisy, but what is worrisome is the trend.  The drop off in oil imports into the US is widening as time goes forward.  

Ponder the impact on US imports of just the crashing production from Cantarell.  To predict that we will see increasing production and increasing imports is to predict what historically has never happened, based on the HL model.

I expect to see $100 oil this year.  The trillion dollar question is what happens to demand.  

Longer term, I think that Simmons is right, i.e., that we will see sustained prices of $200 or more (in 2005 dollars) in 2010.