plucky underdog:

In the late 1950's the world began to really use a lot more oil and gas. The oil finding rate was just about at its maximum, and nationalisation was a storm cloud over the horizon. OPEC hadn't been started It was the good old days of the "free market capitalism" approach to the oil business.

As the world changed from coal to oil for its fossil fuel BTUs, the amount in export trade went up a huge amount. At the beginning of the Second World War, Texas produced 60% of the oil and petrochemicals in world trade, although Texas was producing about 5 million barrels a day.
The rest of the oil in world trade came from Iran and the beginnings of the mid east fields from Shell and Anglo-Iranian Oil, now BP.

When the market was smaller and thinly traded, the decline rate was more closely tied to the performance of individual fields or groups of fields. But as the markets grew, the decline rate had a much larger group of producing field to average it over, so the curve got flattened. At any rate, thats my hypothesis and Im stticking to it. Bob Ebersole

Bob, Rembrandt: your explanations sound coherent... anyone know when the US became a net importer? (the readily-available data series on the Web don't go back as far as 1960, let alone before). I'm going to look at the data for the UK, which recently became an importer as well.

Khebab, westexas: if it (production rate volatility) really is affected by the switchover to export, does this analysis have any interesting implications for the predictive power of ELM? Not trying to muscle in on your zetetic bailiwick, just interested.

It appears that the US became a net importer right around 1949, 21 years before we peaked. Because of a rapid increase in consumption, the US actually showed a pretty rapid collapse in net exports.