In March 1998, Scientific American published “The End of Cheap Oil,” a seminal article by
Colin Campbell and Jean Laherrère that introduced the notion of a near-term peak in
conventional crude oil production. This paper revisits that forecast nearly 30 years later,
comparing it with real-world production data for the period 1998–2023 across four key
geographies: the North Sea, the Commonwealth of Independent States (CIS), the Middle
East, and the world.
The analysis confirms that while the timing of the global peak—around 2006—was broadly
accurate, regional production trends varied significantly. Forecast errors were largely due to
underestimated reserve revisions, conversion factor uncertainties (particularly in Soviet
ABC1 data), and unanticipated discoveries enabled by new exploration technologies.
Using Hubbert Linearization (HL) techniques and updated reserves data or best modeling on
recent production, we present revised estimates of production ultimates for each region. We
find that HL remains a valid tool for post-peak regions and global trends, with an updated
global ultimate of 350 G.m³ (2 200 Gb), compared to 1 800 Gb estimated in 1998.
Far from being invalidated, the 1998 article appears prescient in framing the end of cheap oil
as a structural inflection point. The subsequent rise of expensive and unconventional oil—
subject to the same geological limits—suggests that the debate over resource exhaustion
remains highly relevant. In light of mounting constraints on shale oil and persistent depletion
of conventional reserves, the question posed in 1998 remains open: can we afford to bet that
we will never again face the end of oil?