Shedding Light: Should the U.S. Export Crude Oil and Petroleum Products?

Wednesday, March 5, 2014

Shedding Light: Should the U.S. Export Crude Oil and Petroleum Products?

Need to Know . . .

  • Crude oil exports would not raise the prices of crude oil and petroleum products (e.g., gasoline and diesel)
  • The increase in exports of petroleum products in recent years has not raised domestic prices of gasoline or diesel
  • Had it not been for rising U.S. oil production, the world price of oil would be higher than its current level
  • The push for crude oil exports is a consequence of the fact that not all crude oil is the same. Most of the recent rise in production has been light or “sweet” crude oil while most U.S. refining capacity is configured to handle heavier, “sour” slates of crude oil
  • Exports of petroleum products have benefited the economy by contributing to a reduced trade deficit. Exports of crude oil would do likewise
  • Exports provide incentives for continued development of U.S. energy resources, an activity that benefits manufacturers as well as the overall economy

In the wake of the 1973 Arab oil embargo, Congress enacted a ban on the export of crude oil. The ban was largely academic in that the United States was becoming increasingly reliant on crude oil imports. In 1973, the U.S. imported an average of 3.2 million barrels of crude oil per day (mm b/d) and another 3.0 mm b/d of petroleum products. By 2005, crude oil imports more than tripled to 10.1 mm b/d while petroleum product imports rose to 3.6 mm b/d. Reliance on oil and product imports as a percent of total petroleum consumption increased from 36% in 1973 to 66% in 2005.

Thanks to the hydraulic fracturing and directional drilling revolution, oil production has increased by 2.5 mm b/d since 2008. At the same time, total petroleum consumption in 2013 was 0.5 mm b/d lower than in 2008. As a consequence, crude oil imports declined by 2.1 mm b/d in 2005 to 7.7 mm b/d.1 Total petroleum imports in 2013 were down to 52% of total consumption.

The Energy Information Administration projects that oil production will average 8.5 mm b/d this year and rise to 9.3 mm b/d in 2015, a level close to its historical high of 9.6 mm b/d in 1970. The increased production has led to calls for lifting the ban on oil exports. Some members of Congress are adamantly opposed to allowing exports, expressing fears of rising prices for crude oil and ultimately petroleum products, especially gasoline.

Exports of refined petroleum products are not banned and in recent years, these exports have soared; the U.S. has been a net exporter of petroleum products since November 2010. Nonetheless, members of Congress have attempted to place restrictions on petroleum product exports in the past, and such efforts could be resurrected, especially if product prices ratchet up as they do from time to time for reasons unrelated to exports.2

Why Export Crude Oil?
Given that U.S. crude oil imports total 7.7 mm b/d, one might ask why there is a push to export oil. Why don’t we simply reduce our reliance on imported oil by using the increased production in U.S. refineries, further reducing our reliance on imports?

The reason for allowing exports is that not all oil is the same. Most of increased production in recent years has been in the form of lighter (“sweet”) crude oil. Back when U.S. reliance on imports was much higher, refiners invested billions to reconfigure their refineries to efficiently process heavier crude oil slates because this oil sells at a discount. These refineries can process lighter slate of crude oil, but given the way they have been configured, their efficiency, in terms of the yield of petroleum products such as kerosene, light diesel oil, heating oil, and heavy diesel oil, would fall.

Economically, it makes sense to export the light crude oil that most U.S. refiners do not want to Europe and Asia, where more refineries are configured to handle this product. Light oil sells at a premium compared to heavier oil, and the higher price encourages development of U.S. resources. In the U.S., light oil would have to be discounted given that refiners would see a drop in refining efficiency; the lower price would make the development of some U.S. shale plays less economic.

Increased Oil Production and the Spot Price of Oil
The price of crude oil is determined by supply and demand. On a day-to-day basis, the trend in the price is typically not smooth. Rather, factors such as weather, changes in inventories, changes in the value of the dollar, new economic reports signaling an upturn or downturn in the economy, terrorist actions halting production, and changes in the amount of excess worldwide production capacity cause fluctuations around the long-term trend in oil prices.

Daily spot oil prices reflect the confluence of all these factors. Light crude oil is currently selling at spot prices below the world price of oil. As shown in Figure 1, the spot price of West Texas Intermediate (WTI) is lower than the Brent spot price. The latter is a good indicator of the cost of incremental (imported) oil supplies. In addition, gasoline prices are tracking the Brent spot more closely than they do the WTI spot price. This was not always the case; prior to 2010, the long-term difference between these two spot prices averaged 95 cents per barrel.

Figure 1 – WTI and Brent Spot Prices

Source(s): U.S. Energy Information Administration

There are two main reasons the WTI spot price is below the Brent spot price. First, because of the surge in oil production, the ability to ship oil coming down from North Dakota is hindered by the lack of pipeline capacity. Oil must therefore be shipped by rail, which is more expensive; to compete, that oil must sell at a discount. Pipeline capacity is increasing, however, making it easier to ship oil through the Midwest. As a result, the WTI spot price is rising and the gap between it and the Brent spot price is narrowing.

The second reason is that refiners are not willing to pay a premium for lighter oil. To compete, the lighter oil that is being produced in places such as North Dakota must be priced competitively with less expensive, heavier crude oil because refiners can process the latter more efficiently.

Crude Oil Exports and the Price of Crude Oil
Critics argue that exports would raise the domestic price of oil and therefore the cost of petroleum products such as gasoline and diesel fuel as well. In reality, any impact on domestic petroleum prices would be minimal because crude oil prices are determined in the world market. If the U.S. maintains its ban, the domestic price of oil paid by some refiners might be lower, but the prices paid by other refiners would reflect the world price because they would still rely on imported oil. It is the incremental source of oil that sets the price in the overall market.3

There is a touch of irony in the argument that allowing exports would raise the price of oil because others have complained that the significant increase in U.S. oil production in recent years has not reduced the price. In fact, the 2.5 mm b/d increase in U.S. production since 2005 has contributed to an easing of the pressure on the world price of oil.

Had U.S. oil production not risen, the world price would be higher than it is today because the world demand for oil continues to grow. One way to approximate the impact this increased production has had on the world price is to ask how much higher the average price of crude oil might have been in 2013 had U.S. production not risen. The short-term impact can be estimated using the following relationship linking oil supplies to the price of oil:

% Change in Price = % Change in Supply / Elasticity of Demand

Most estimates of the short-term price elasticity of demand are between -0.05 and -0.10. The short-term price elasticity of oil is low because of limited substitution possibilities. A 2.5 mm b/d reduction in the supply of oil in 2013 would have been equivalent to a 2.8% decline in world oil supplies. Given the range of short-term oil price elasticities, the relationship shown above indicates that a 2.8% decline in supply would have raised the world price by 28-56%. Given that the refiner acquisition cost of imported oil averaged $98.50 in 2013, the price of oil would have been between $126 and $154 per barrel, at least in the short term.4 Such an increase would have added $0.65-$1.32 per gallon to the cost of producing petroleum products.

Petroleum Product Exports
Petroleum product exports have increased sharply since 2005 and the U.S. is now one of the largest gross exporters of these products (Figure 2). U.S. exports of petroleum products and crude oil are four times their level in January 2005, when exports averaged 0.92 mm b/d. In October 2013, exports averaged 4.0 mm b/d. Most of these exports (97% in October 2013) were petroleum products, not crude oil. The decline in petroleum consumption in the U.S. since 2005 coupled with the surprising increase in oil production has resulted in a large supply of petroleum products available for export.

Figure 2 – U.S. Exports of Crude Oil and Petroleum Products

Source(s): U.S. Energy Information Administration

Petroleum Product Exports and Product Prices
While concerns have been raised about oil exports, there has been no move to limit the exports of petroleum products. Still, one might think that the rapid rise in product exports might have increased the domestic prices of the products that were exported.

As shown in Figure 3, the prices of gasoline and No. 2 diesel fuel closely track the refiner acquisition cost of crude oil. Unlike spot prices, the latter price is a measure of the average price refiners pay for all the oil they purchase, domestic and imported. While movements in spot prices are closely followed in the press, most oil is sold under contract, not in spot markets.

Figure 3 – Refiner Acquisition Cost and the Prices of Gasoline and No. 2 Diesel

Source(s): U.S. Energy Information Administration

Figure 3 provides a quick overview of the relationship between the price of crude oil and product prices. It is obvious that the two sets of prices move closely together. An econometric test of the relationship between product prices and crude oil provides even more compelling evidence on the link between oil and product prices. A regression equation in which product prices are a function of the price of oil found that from 1995 through the end of 2013, these prices moved in lockstep.

As shown in Figure 4, the U.S. became a net exporter of petroleum products, mainly diesel fuel, in November 2010. It is possible to test whether the switch to being a net exporter of petroleum products affected the relationship between oil and product prices.

Figure 4 – Net Petroleum Product Exports

Source(s): U.S. Energy Information Administration and MAPI

A statistical test5 lends support to the view that there was a break in the relationship between crude oil and product prices after 2011. The difference, however, is somewhat surprising because the price of diesel fuel rose less rapidly than the price of crude price in the period 2011-2013 than it did during 1995-2010. This suggests that petroleum exports have not raised domestic petroleum prices.

The difference between product prices and the price of crude oil expressed as a percent of the price of crude oil is one measure of the margin between the two prices. The trends in the gasoline and diesel fuel margins are shown in Figure 5. These margins have trended down over time and have been relatively flat since October 2009. The fact that they did not rise as product exports took off provides further evidence that petroleum exports have not affected domestic prices.

Figure 5 – Retail Gasoline and No. 2 Diesel Prices as a Percent of the Refiner Acquisition Cost of Oil

Source(s): U.S. Energy Information Administration

Benefits of Petroleum Product and Oil Exports
The questions about the wisdom of allowing oil exports can be raised with respect to exports of corn and myriad other agricultural products that are major U.S. exports; for these products, the U.S. is not dependent on imports. If exports of the latter were prohibited, their prices would fall because the U.S. does not rely on imports, at least temporarily. If export bans are justified and imposed because of their potential impact on domestic prices of the goods that are exported, U.S. exports would fall dramatically. Farmers would eventually cut back on production in response to lower prices and the prices of many agricultural products would then rise.

Again, exports of oil will not lead to increased prices because oil is traded in the world market and because the U.S. is a net importer of oil. Still, one might ask if there are any positive benefits from exporting crude oil. One obvious upside is the impact on the U.S. trade deficit. In 2013, the current account deficit was $502 billion—better than it was in 2005, when it stood at $716 billion. As shown in Figure 6, the value of petroleum exports increased from $22.7 billion in 2005 to $134.3 billion in 2013, or by $111.6 billion. The increase in the value of petroleum exports is equivalent to 52% of the reduction in the current deficit between 2005 and 2013.

Figure 6 – Value of Petroleum Exports

Source(s): U.S. Census Bureau and MAPI

If the ban were lifted, oil exports would generate $33 billion per year in revenue at today’s prices if an additional 1.0 mm b/d were exported. Given that oil production is expected to increase by 1 mm b/d in 2014 and another 0.8 mm b/d in 2015, and that most of this will be light oil, the ability to export 1.0 mm b/d is feasible.

The case for allowing exports of crude does not rest solely on the fact that exports would reduce (but not eliminate) our balance of trade problem. To the extent the ban on crude oil exports keeps domestic crude prices depressed, development of U.S. energy resources is discouraged. This development spurs economic activity (including activity in the manufacturing sector) and creates jobs. The ban on exports of crude oil imposes a real cost on the economy.


(PA-136 PowerPoint slides here for members only. Login required.)

  • 1. Crude oil imports did not decline by as much as production increased and consumption declined because some of the imported oil was used by refineries for their petroleum exports, which have increased rapidly in recent years.
  • 2. In 2005, the Stop Heating Oil Exports Act of 2005 was introduced in the Senate after a dramatic spike in the spot price of oil. The bill would have authorized the secretary of energy to temporarily prohibit the exportation of finished petroleum products; it went nowhere, perhaps because by the time it was introduced the spot price was rapidly falling.
  • 3. This proposition comes from basic economic principles and was recognized by the Department of Justice long ago in testimony given in regard to how the price of oil is determined in downstream (end) markets. See the testimony of John H. Shenefield, assistant attorney general, antitrust division, Department of Justice before the Subcommittee on the Judiciary, United States Senate “Concerning Oil Company Ownership of Pipelines,” June 28, 1978.
  • 4. The situation could have been worse had it not been for the development of hydraulic fracturing and directional drilling; prior to the introduction of these technologies, U.S. oil production was declining. In all likelihood, that decline would have continued and the reduction in world oil supplies would have exceeded 2.7 mm b/d. Some might argue that OPEC surplus production capacity could have made up for that loss. In 2013, however, OPEC surplus capacity was, by historical standards, at a very low level and it is not obvious that OPEC (really, Saudi Arabia) could simply have turned on the tap to prevent prices from rising. Even if it could have, the ability of OPEC to meet increasing world oil consumption would at some point diminish. At that point, the world price of oil would rise rapidly in the absence of increased U.S. oil production.
  • 5. The Chow test was used to determine whether the coefficients of a regression analysis linking petroleum product prices to the price of oil for the period 1995-2013 are statistically different from coefficients estimated over the sub-periods 1995-2010 and 2011-2013.