Just in Time to blame for gasoline price spikes

I think I first saw someone (Nader?) making the title's claim on the Tonight Show. With Johnny Carson.

It has been frequently charged that the oil industry has been influenced by JIT (see here (2004), here (1996), and here (1996)). The argument goes like this: The number of refineries in the United States has fallen from 279 in 1975 to 205 in 1990 and further to 149 in 2004. As a result, the industry is susceptible to supply shocks, which causes spikes in prices and subsequently reduction in domestic output. The effects of hurricanes Katrina in Rita are given as an example: in 2005, Katrina caused the shutdown of 9 refineries in Louisiana and 6 more in Mississipi, and a large number of oil production and transfer facilities, resulting in the loss of 20% of the US domestic output. Rita subsequently shut down refineries in Texas, further reducing output. The GDP figures for the third and fourth quarters showed a slowdown from 3.5% to 1.2% growth. Similar arguments were made in earlier crises.

However, JIT students question whether JIT as it has been developed by Ohno, Goldratt, and others was even in use by the petroleum industry. JIT requires a reduction in inventory capacity, not production capacity. From 1975 to 1990 to 2005, the annual average stocks of gasoline have fallen by only 8.5% from 228,331 to 222,903 bbls to 208,986 (EIA data here). Stocks fluctuate seasonally by as much as 20,000 bbls. During the 2005 hurricane season, stocks never fell below 194,000 thousand bbls, while the low for the period 1990 to 2006 was 187,017 thousand bbls in 1997. This shows that while industry storage capacity has decreased in the last 30 years, it hasn't been decimated.

On the other hand, the storage capacity as a fraction of the daily use has decreased. American consumption has increased fairly steadily during the same period that the storage capacity has been slightly diminished. In 1975, there were approximately 20 days of storage (stocks divided by domestic production); in 1990, 14 days; in 2005, 11.5 days. When expressed as a fraction - or as the number of days' use - it has declined. This is an industry-wide phenomenon, however, and not part of a corporate plan to reduce inventory as JIT would require.

During the same period, natural gas transportation was deregulated. There has been a gradual shift away from fuel oil toward the use of natural gas for heating. It is therefore not surprising that fuel oil stocks have declined from 92,000 thousand bbls to 37,000 thousand bbls, with most of that drop coming in 1979-1986 (to 46,000 thousand bbls). EIA data in file pet_stoc_wstk_dcu_nus_m.xls at the EIA website.

Further, domestic gasoline production capacity has been increased since 1975 from 14,961 thousand bbls per day to 15,572 in 1990, and 16,894 in 2005 (EIA data). In addition to the crude that is imported for production of oil, refined gasoline importation has increased from 200 thousand bbls/day in 1982 to 1,110 thousand bbls per day in 2006, or about 6% of the total. Most of that oil is imported through the Eastern PADS district, not the Gulf Coast. Thus, the domestic gasoline production capacity has not only increased from 1975, but the total import and domestic capacity has increased while the natural variation risk has been distributed (at an increased exposure to political risk for the imported fuels).

At the same time that natural gas was deregulated (thanks, Jimmy Carter!), airlines were deregulated. There has been increased refinery output of kerosene (jet fuel) as a result. From 1982 to 1990, jet fuel production increased from 753 thousand bbls per day to 1,311 thousand bbls per day, and then to 1,557 thousand bbls per day in 1999, where it peaked as a result of the subsequent economic slowdown. It currently stands at 1,538 thousand bbls per day average over the past year. Thus, it seems amazing that gasoline production has increased at the same time as jet fuel production with fewer refineries. This is an increase in effective production capacity, not a decrease.

Other questions would have to be addressed in order to determine whether the 2005 fluctuations were due to adoption of Just In Time, other industry practices, or other factors:

• Has domestic production capacity grown increasingly concentrated in the Gulf Coast region since 1975? Why or why not? It seems likely that Gulf of Mexico oil production has increased since then, and that production off the California coast has declined in the same period, so my estimate is that the answer to this question is yes. If my intuition and the reason for it are both correct, this is a change caused by something other than JIT.
  
• Has domestic storage capacity grown increasingly concentrated in the Gulf Coast region since 1975? Why or why not? It seems likely that storage follows production, so again I estimate the answer is yes, and again has little to do with JIT.
  
• Were there any periods prior to 1975 in which a hurricane hit the Gulf Coast, did prices fluctuate as a result, by how much, and what were the conditions immediately preceding (i.e., were spot and futures prices already rising or falling)? I believe that a large part of the problem in 2005 was that the market was already nervously watching the increase in prices when the storms hit.
  
• Have the reductions in numbers of refineries been actual shut-downs of plants, or consolidations of multiple plants under consolidated management?
• The automobile industry is arguably more consolidated than the oil industry, though both are clearly engaged in strong competition. Yet, only one or two automobile manufacturers are actually practicing or attempting to practice JIT, while the others have made starts toward it and then abandoned the quest (see, for example, GM's experience with NUMMI, or Chrysler's demise since the Daimler-Benz takeover merger. It seems unlikely, therefore, that the entire oil industry could have adopted JIT principles in unison in the 1970s and followed through until 2006. Is there any claim on the part of major refiners or distributors to be using JIT methods?
  
• What was the effect of growing worldwide -- especially Chinese -- demand for oil? In other periods, were the markets in a state of contango or backwardation? For example, gasoline stocks were lowest during the 1997 period, probably because the price of oil fell so low that nobody wanted to store product while prices were falling.
• How much of the reduction can be explained by the unpredictability introduced by the multitude of regulations introduced between 1989 and 1992? In that period, the oxygenated fuels legislation was passed, giving EPA the power to demand sudden changes in fuel blends (so-called "boutique" fuels). It makes no sense to keep finished gasoline on hand when the EPA could suddenly declare a new blend; thus there has been a rise in stocks of blending distillates concurrent with the decrease in finished gasoline.

Economic vs. JIT rationales

Oil and other companies routinely shut down facilities for reasons other than the application of JIT. One of those reasons may be economic rationalization: when the benefits of operating no longer outweigh the costs, including opportunity costs, the plant may be economically inefficient. JIT has never subscribed to such considerations directly; in fact, following Waddel (Rebirth of American Industry) this thinking may be based on Brown-style accounting and Sloan management, which are based on DuPont's use of ROI as the measure of profitability. Waddel contrasts that with the cash flow definition of profitability used by Henry Ford and Toyota. Ford during the Rouge era would have sustained and Toyota does sustain practices and plants that appear to be economically inefficient if they serve a strategic interest, especially if they improve flow and speed. Thus, another problem with JIT is that some of the application appears to be counterintuitive and at odds with traditional accounting (GAPP) and managerial economics as practiced in large corporations post-Sloan.

The real reason

In my research, I have found nothing that indicates that refiners are actively attempting to apply JIT theory to their operations. The real reason for the price spikes in 2005 were:

• increased demand for petroleum in China and elsewhere,
• inability or lack of desire on the part of the Saudis and other OPEC members to increase supply
  
• nervous markets as a result of trouble in Iraq, Argentina, and elsewhere,
• a recovering economy in the US,
• and two large storms that hit the Gulf Coast nearly simultaneously.

George Bush was right when he said that we are addicted to oil. We refuse to do anything to change our lifestyles -- hey, we can quit anytime! Instead, we blame the oil companies (who bring us what we want), the Arabs (who are actually a minor supplier compared to domestic wells, Canada, Mexico, and Argentina), George Bush and Dick Cheney (as if the problem started in 2001), and anyone else who could plausibly be blamed. But price is a result of both demand and supply, and we as consumers are responsible for one side of that equation.

We have met the enemy, and he is us.

The Accounting Chains on American Industry

I just finished Bill Wadell and Norman Bodek's Rebirth of American Industry. As usual, I feel I have to get the negative parts out of the way. First, the title is misleading, since the book isn't really a story about how American Industry is bouncing back, but rather about what needs to happen in order to be reborn. Second, the book is not aesthetically appealing, with text that goes too far into the spine, low quality illustrations (at least one figure looks like it was printed on a dot matrix printer), and sidebars that would have worked much better in a larger format. Third, the book has some strange editing errors, including inexplicable line breaks and the like. I didn't pay for it, so I cannot say whether it merited the $47 cover price. In fact, after a cursory glance, I decided I wasn't interested.

[NOTE: I was reading a preliminary, pre-production edition, so the aesthetics will likely be different than what will be for sale.]

After hearing my wife make a "hmm.." noise a few times while she was reading it, I decided it might be worthwhile, so I finally read it. I'm glad I did. Although the authors claim it is not a history book, it does in fact shed light on a very important aspect of American corporate history: the imposition of the DuPont definition of profit, the Sloan management method, and the Brown accounting method onto American industry. In fact, Sloan and Brown were picked by DuPont and their methods reflect his definition of profitability, so all the evil essentially flowed from one man. At the end of World War II, with Ford Motor Company having faltered as Henry lost his mind, the DuPont/Sloan/Brown system stood at the pinnacle of the most important industry in the only country in the world left relatively unscathed by the war. Their system was adopted by every other American business for this bit of luck, not because of any inherent merit.

The Sloan system (the common, short-form name) essentially transformed manufacturers into marketing companies with a manufacturing function. The authors argue that the definition of profit is the most important aspect to consider: for DuPont, profit meant Return on Investment (ROI). The accounting system that counted inventory as an asset and people as a cost followed from that principle. No wonder GM treated its workers and suppliers (and suppliers' workers) so poorly, no wonder the UAW felt so strongly, and no wonder they are both so screwed up today. In fact, you can clearly see why GM would push problems (what most would call business risk and employees) out the door to suppliers. Once you have decided on that strategy, you have declared that you are no longer a manufacturer with a sales office, you are a marketing firm for product manufactured by someone (you are indifferent as to who manufactures).

On the other side of Pacific, and previously at Ford on this side, the definition of profit had been cash flow. In Ford's terms, it was when you had more money in the safe at the end of the week than the beginning. In Ohno's terms, the Toyota system was to try to cut the time between receipt of order and receipt of pay for that order. In such a system, inventory is no asset, but well-trained people and process speed (and therefore quality) are.

To put it in terms of the SIPOC model, Suppliers contribute Inputs to the Process. If you want that process to be fast, the inputs better be good. That means that the Suppliers better be good and getting better. Customer focus means speed implies both profit and quality. That's win-win-win thinking (investors, employees and suppliers, customers).

I don't agree with Bill about the necessity of manufacturing or the balance with service. At the beginning of last century, America was mostly an agricultural and natural resource extraction economy, yet Americans were relatively prosperous. In the middle of the century, the balance shifted toward manufacturing, and Americans were more prosperous. At the beginning of this century, the balance has shifted toward services, and Americans are even more prosperous. People tend to forget that the "services" industries include research scientists, chemists, engineers, architects, lawyers, financiers, doctors, nurses, and so on. These aren't low-pay jobs, and they all contribute to higher productivity and standards of living by people who are engaged in the traditional non-service sectors: construction, extraction, agriculture, and manufacture.

However, I do agree with Bill and Norman about the problem of trying to become lean with a Brownian accounting system. That, like the "management by numbers" stupidity of Sloan "management", serves only the near-sighted and non-manufacturing-minded DuPont definition of profit. Deming was right: the idea that someone could graduate with an MBA and step into the front office of a manufacturer without ever having spent time on the factory floor is a fraud. It was a fraud allowed to flourish in a time when Sloan companies were competing against other Sloan companies: marketing vs. marketing. Once a Manufacturer stepped into the ring, they were doomed to fail. Unless they change their definition of profitability and the management and accounting systems that support that, the Sloan companies are not going to compete with a Toyota.

I'm not an accountant and I don't know much about GM or Sloan. This book was exciting to me because of an interest I have had lately: the question posed by the Austrian Economists blog, "Does Management Science have anything to say that Economics doesn't know?" The common answer is that they don't; that this would be like physicist learning something from engineering, which is applied physics. I disagree: physicists are looking at impersonal, inanimate objects and forces, while economists are looking at human beings and their choices. A corporation isn't a deterministic black box; it is filled with non-deterministic beings trying to figure out the best methods for uncertain and perhaps conflicting ends. The economist may assume that corporations exist to generate profits for the owners, and everything follows from that. Shouldn't the economist consider how the corporation defines "profit"? Because it appears to matter.