By Moshe Adler
Q: To which of the following debacles is the rollout of Obamacare most similar? (a) The several-year delay in the rollout of the Boeing 787? (b) The failure of Halliburton to deliver food and mail to our fighting soldiers in Iraq in spite of exorbitant payments? (c) The theft from New York City of $500 million by the contractor that was hired to create CityTime, a computerized payroll system? If your answer is (d) all the above, you got it right.
Each of these examples involves a large investment in a product for a single client. To use the economists’ terminology, each involves “specific capital.” The best way to produce specific capital, economists Benjamin Klein, Robert Crawford and Armen Alchian explain, is to produce it in-house. Why? Because when specific capital is involved, outsourcing means a total loss of control over the product under conditions that are hard to reverse.
In the case of the 787 Dreamliner, the delays were caused by Vought, a South Carolina firm that had been contracted by Boeing to produce the fuselage. Not only was the product that Vought was hired to produce specific to Boeing, it was also new—Boeing had never before produced a passenger fuselage made of composites—which meant that Boeing’s specifications for the product kept changing. And just as any management manual could have predicted, Vought failed repeatedly to produce a fuselage that met Boeing’s needs. But in spite of all the grief, Boeing could not simply replace its supplier, because starting from scratch would have delayed production even longer. So, for more than two years, Boeing tried to get Vought to deliver. In the end, Boeing had no choice but to buy the company and bring fuselage production in-house.
Of course, not all of the parts that go into a Boeing plane need to be produced by Boeing. The company has never produced engines, for example. But engines are not specific to Boeing and are produced by several manufacturers. If one had produced a faulty product, it could have been replaced relatively quickly. The key, then, is the level of specificity. The product that Boeing was buying from Vought was highly specific to Boeing, and Boeing therefore ultimately had to be directly in control of its production.
Boeing’s management should have known this, because it is part of the basic instruction in any outsourcing manual. Take, for instance, the privatization guidance package distributed by the World Bank. The bank promotes privatization all over the world, but it warns that for a government to be able to control its contractors and know what an appropriate payment for their services is, it must also produce the same product itself. Duplicate production is always difficult when a specific product is involved, but it is simply impossible when that product is new and being developed for the first time.
When it bought Vought, Boeing paid a huge sum for a company that had no other bidder and a bad track record. Perhaps management had something to hide. Boeing’s unions had warned about the pitfalls of outsourcing fuselage production, and to avoid an attention-grabbing hostile takeover fight with the owners of Vought, apparently even $1 billion of shareholders’ money was not too high a price to pay.
Unfortunately for them, soldiers are not like airlines. Had the 787 been a bad plane, the airlines could have switched to Airbus. But where could soldiers who were defrauded by Halliburton go? KBR, a Halliburton subsidiary that had a contract to deliver meals and mail to our fighting soldiers in Iraq, didn’t deliver either. No conclusions were drawn from this failure, however, despite the fact that from 1776 until the war in the Balkans, when the Clinton-Gore team “reinvented government,” cooking and mail delivery for soldiers were provided by soldiers. So our fighting soldiers went without mail and without cooked food. But instead of returning to a system that was both efficient and cheap, KBR, the biggest nonunion construction company in the U.S., continues to be a favorite contractor of the military, receiving a $500 million no-bid contract in 2010 and an additional $250 million contract in 2011.
Even local governments have their boondoggles. Mayor Michael Bloomberg’s scheme to modernize New York’s payroll system, CityTime, ended up billing the city $700 million. Fortunately for New York residents, the union representing public employees was tenacious in working to expose the corruption involved in the contract for CityTime, in spite of strong resistance by the mayor. In the end, the U.S. attorney for Manhattan stepped in and the contractor agreed to pay back $500 million. Two hundred million dollars is still almost three times the $73 million original cost of the contract, and the cavalier overpayment of sums that constitute an almost tenfold cost increase make it clear that government officials cannot be trusted with private contracting.
The Obama administration has now hired yet another contractor to fix the problems that the original contractors of the Affordable Care Act software created. This is the wrong solution. The government should hire its own programmers to do the job. When it comes to a product that is truly unique and has only one client, history has shown that there is no alternative to production in-house.