Peak Oil is Here

I will write a follow-up post asking what I got wrong and what (if anything) I got right with my analysis below.
When done, I'll put in a cross-link.

Discussions I've seen of peak oil focus upon it as an issue in geology: how many years until production begins to decline [name your fossil fuel]. As an economist, however, that's irrelevant, reflecting a misperception that underlies the 1980 Simon - Ehrlich bet. In an economy, there are many margins of adjustment. As a starting point what matters is rather the time path of real prices. Since this blog focuses on the auto industry, the second margin is fuel efficiency of vehicles. I won't get to the second piece for several paragraphs, and then only briefly.

The market price of a depletable resource need not follow any particular time path – many (silicon) are abundant beyond any likely need, and so the price reflects the costs of extraction and short-run shifts in demand and installed (extractive) capacity. Over time, however, we've extracted the most readily accesible reserves of petroleum, and "scarce" or not, price will rise. (Likewise, we're seeing an increase in demand, given the success of East and Southeast Asia – and above all China – in lifting much of their populations out of the grimmest sort of poverty.

The early years of the industry saw many technical improvements in extracting petroleum and refining and transporting it as gasoline. Combined with discoveries in the Persian Gulf, prices were relatively stable (and low) from the end of World War II (1946) through the early 2000s, albeit with a spike around 1981. Indeed, in the 1990s energy prices were at their lowest in human history, as per an earlier blog post, "Are Gas Prices High?," particularly if we take into account the rise in real (US) incomes.

On impulse I looked at prices for the first time in three years. As a devil's advocate let me suggest that we are now seeing peak oil, with prices rising and (my hypothesis) not falling here on out. Here are 3 graphs. First, here are import prices of petroleum in 3 large economies. Now my first caveat is that to Germany and Japan what matters is the price in (respectively) the Euro and the Yen, but during this time frame forex rates have not swung by even two-fold. Looking at post-1949 domestic U.S. prices gives the same story; as per the first graph, I've shaded in the 1990s to emphasize that prices were below the post-WWII trend through 2000, and significantly below the post-war trend when extended through early 2014.

But as an economist peak oil is about price trends, not levels. So for the final graph I calculated the price trend over 1976-2000 and compared it with the trend over 2001-2014 (I truncated the trends berlow where they intersect). In the first period, prices fell at 6.3¢ per year; in the second, they rose at an annual 15.4¢ rate. Now many things could affect this. In the short run, the Chinese economy may see a recession. The timing of when new production comes online is potentially lumpy, and can depress prices until global growth catches up. In the longer run, high prices encourage the development of alternatives; natural gas in the next decade, solar power thereafter. In other words, this is informal analysis, and is not embedded in a model that spells out alternatives in terms of rates of change on the demand and the supply side.

For the auto industry, there are several margins of adjustment. First, there's an improvement in fuel efficiency; my '88 Chevy pickup is lucky to get 10 mpg, my Chevy Cruze gets 35 mpg running up (and down) hills to get into town, and over a 25 mil stretch of highway driving I've peaked at 50 mpg. So in the short-run I've been able to offset the rise in prices at the pump – until I need to haul a load of rock dust. But if the trend continues, then the next 3 years will see prices up by roughly a half dollar, or about 14%. As a PACE judge the technologies I see suppliers bringing to market suggest that the industry can offset that magnitude of price increases for another decade. Thereafter alternative fuels and shifting driving habits provide additional margins of adjustment – in the normal week neither my wife nor I drive more than 75 miles in a day, well within electric vehicle range. My old house sits 50 feet from the regional natural gas pipeline. While not yet fully cost competitive (and then there's the base vs peak load issue), solar power is not a depletable resource, and natural gas is depletable, further from "peak" (and potentially renewable as a biofuel).

The underlying issue is the extent to which the industry should place bets of future high petroleum prices. Public policy, particularly in Europe, has already decided that. Whether (or rather at what rate) we will buy such vehicles – and adjust our driving habits – will be a function of prices. From the industry's perspective I am however not particularly worried about our running out of oil, even while believing that we're seeing peak oil.