My guess is that the production drop due to reduced demand will be from October forward, but in any case if we take the 2008 EIA data (subject to revision, frequently downward) at face value, the 2005, 2006, 2007 and 2008 annualized production volumes are as follows (C+C, using 365 days for 2008):

2005: 26.9 Gb
2006: 26.8
2007: 26.6
2008: 27.0*

*Based on preliminary data through August

If we round off to the nearest billion barrels, it's four years at 27 Gb per year. Alternatively, the cumulative shortfall between what we would have produced at the 2005 rate and what we actually produced would be 300 mb, using 27 Gb for 2008.

Of course, different crude oils trade at different prices, but if we use Brent as an index price for comparison purposes, the total dollars paid for crude would like like this (assuming $100 for 2008):

2005: $1.48 Trillion ($55 per barrel)
2006: 1.74 ($65)
2007: 1.92 ($72)
2008; 2.7 ($100)

This pattern of flat to declining production versus rising oil prices is exactly what we saw in the initial Texas and North Sea declines, two regions developed by private companies, using the best available technology, with virtually no restrictions on drilling:

http://www.theoildrum.com/files/TexasAndNorthSea.png

If we use 27 Gb for 2008, we would have increased annual production by 100 mb (per year) relative to 2005, so we paid (for comparison purposes) $1.2 trillion more for an incremental increase of 100 mb, or about $12,000 per barrel of incremental production.

However, the reason for using cumulative production is because that directly relates to the area under the production rate versus time curve, which is the argument that Hubbert made back in 1956. The bottom line is that the crude data show a cumulative shortfall relative to 2005, even with the contribution from unconventional production.