Michael Hiltzik has a fantastic column on Boeing’s outsourcing disasters in the LA Times; it’s well worth reading the whole thing, complete with a link to a prescient 2001 paper by Boeing technical fellow LJ Hart-Smith.
Hiltzik’s point, which is undeniable, is that Boeing’s outsourcing mania has cost it billions. It’s not a new idea (Reuters ran this special report in January), but it’s rarely been this well expressed:
Boeing’s goal, it seems, was to convert its storied aircraft factory near Seattle to a mere assembly plant, bolting together modules designed and produced elsewhere as though from kits.
The drawbacks of this approach emerged early. Some of the pieces manufactured by far-flung suppliers didn’t fit together. Some subcontractors couldn’t meet their output quotas…
Rather than follow its old model of providing parts subcontractors with detailed blueprints created at home, Boeing gave suppliers less detailed specifications and required them to create their own blueprints.
Some then farmed out their engineering to their own subcontractors. At least one major supplier didn’t even have an engineering department when it won its contract.
Not only was all this forseeable, it was foreseen — not only by the unions, but also by executives. And, of course, the aforementioned Hart-Smith:
Among the least profitable jobs in aircraft manufacturing, he pointed out, is final assembly — the job Boeing proposed to retain. But its subcontractors would benefit from free technical assistance from Boeing if they ran into problems, and would hang on to the highly profitable business of producing spare parts over the decades-long life of the aircraft. Their work would be almost risk-free, Hart-Smith observed, because if they ran into really insuperable problems they would simply be bought out by Boeing.
What do you know? In 2009, Boeing spent about $1 billion in cash and credit to take over the underperforming fuselage manufacturing plant of Vought Aircraft Industries, which had contributed to the years of delays.
The lesson here is that Boeing executives, just like most of the rest of corporate and political America, were incredibly bad at pricing moral hazard and tail risk. Outsourcing is a bit like taking collateral from your repo operation and investing it in subprime credit. Most of the time, you make a small amount of money — and then, occasionally and unpredictably, you lose an absolute fortune. Boeing was picking up pennies in front of a steamroller, and ended up getting crushed.
I do wonder what proportion of corporate “efficiencies” are false ones along these lines. Did Mark Hurd improve HP’s margins by cutting back on R&D expenditure? Or did he sign the company’s long-term death warrant? And of course when Win Neuger’s reach for yield in the AIG securities-lending operation was truly disastrous.
Hiltzik concludes:
The company now recognizes that “we need to know how to do every major system on the airplane better than our suppliers do.”
One would have thought that the management of the world’s leading aircraft manufacturer would know that going in, before handing over millions of dollars of work to companies that couldn’t turn out a Tab A that fit reliably into Slot A. On-the-job training for senior executives, it seems, can be very expensive.
The sad thing is that this lesson has to be learned the hard way so many times. Can’t anybody else learn from Boeing’s mistakes?
The underlying problem with the 787 is evident every time I read about the problems with the aircraft’s “supply chain.” This is not a supply chain problem, but rather a design chain problem. Boeing seems to not understand this difference.
Large aircraft are supplied in sections, and this is possible because each section is modular with highly specified interfaces. However, the design of the sections is not something that can be modularized. The design of each section is highly interdependent with the other parts of the aircraft. In fact, there may be no other product whose parts/sections are as highly interdependent than a large aircraft…a slight change in the center of gravity of one section can have large impacts on the other sections…a slight change in the surface geometry of one section can drastically change the airflow over other sections.
I even recall at one point a few years back that Boeing thought that disintegrating the design would actually shorten the development time…a deeper understanding of the development process would have acknowledged the increased time needed to design an integrated aircraft across companies, countries, cultures, and languages, whereas previously the aircraft and the production process were designed by internal teams.