Open Access Factories
This semester, I have been teaching a seminar simply called "Bailouts." This week, we have been talking about the automobile industry. One of my students, Aaron Moshiashwili, put forth an interesting idea in his written work for the week. In the seminar, I have stressed that the idea is not to save a particular company but the productive assets that company represents--a point that generalizes to many other contexts in corporate law. In other words, we shouldn't care about the logo that is on the door, but we should care about what goes on inside the building. Regardless of whether they make it or not, the automobile companies are going to create a lot of excess capacity in physical plant and human capital.
Moshiashwili had the following thought:
What if a similar provision were implemented for the auto industry? Right now, the Big Three have an AWFUL lot of resources going to waste. And, there are several tech startups right now that are TRYING to make new tech cars, but are finding, surprisingly, that making a car is a fairly difficult endeavor.
So what if one of the requirements of the bailout is that the Big Three, rather than reducing their capacity, are required instead to provide resources to startup auto makers at a flat/reasonable fee? Rather than laying off engineers and closing plants, someone like Tesla motors could pay Ford to have Ford engineers work on the parts of the car Tesla was having a hard time with, and then retool a plant to manufacture Tesla cars. (Tesla makes electric sports cars; they've got an advanced electric motor and drive train, but Tesla had a very hard time getting the gearing working. That strikes me as something that Ford would probably be quite good at.) Right now Tesla Roadsters are being made at a Lotus plant in Wales, while American companies are near to closing shop. Tesla is an American company. This seems like a dumb situation.
I'm not claiming that this would be a panacea. Tesla has managed to make slightly more than 200 cars in two years of production; hardly enough to keep GM in business. But the thing is, once resources exist, people learn how to use them. Right now, if I have a brilliant idea for a new car, what can I do? Starting a new car company is five or ten orders of magnitude more complex than most start-ups. But if I know that all I need is a design and financing, after which I'll be able to pay someone who already owns a factory to help me make it (rather than making a deal which might require giving up the rights to my design, or having to make my own factory) I'm much more likely to give that idea a go, rather than tossing it in the bin.
I understand Moshiashwili to advocate a mandatory system--Ford would have to accept Tesla's work. (Ford is not one currently part of the automobile industry bailout, but the point here is general and not specific to Ford.) The idea is protean to be sure, but it's a clever thought that I had not seen anyone mention. The devil is in the details, and maybe there are too many devils in these details to make it a workable solution. For example, how would you prevent an automobile company from deliberately disqualifying itself from such a program and by sabotaging its excess capacity through layoffs or other devices? Is it reasonable to require an automobile company to keep such excess capacity "online" just in case an entrepreneur emerges who wants to take advantage of plants and human capital? If it's such a great idea, then why don't the automobile companies do it voluntarily for pay?
It certainly seemed like an idea that should be part of the discussion.
Prof. Lawless,
Really? Mr. Moshiashwili is certainly being creative, but the analogy is terrible. Unbundling network elements (such as the last-mile of copper loop to a residential neighborhood) works (at least in theory) because the network elements are commodities -- that is, a piece of copper wire is a piece of copper wire, and whether AT&T or SBC or Iowa Telecom laid that wire doesn't change what's there. Unbundling an element for a competitor was also extremely simple: just flip a switch and the loop is connected to a competitor's circuit.
In contrast, GM's and Ford's factories and engineers are highly specialized: Bobby Jones is a human being with ten years of experience in designing drive trains for jeeps; Sally May works in biology before turning to work on hybrid gasoline-electric engines. People aren't commodities, so why on earth would we think it possible that Tesla could force GM to rent it an engineer that would actually help it's projects? And it's a lot more difficult to add or subtract people from a project than commodities -- how would Tesla even know which GM engineers it would want to hire?
What is worse, is that any type of unbundling solution to the current problem is totally unnecessary. If Tesla really wants all those GM and Ford engineers, let the companies go belly up and Tesla will hire them. Moreover, it's not like Tesla has no other choices. In the telecom world, local regulations and high fixed costs make it extremely costly for a new competitor to string a wire to already-served customers. In the world of making automobiles, Tesla can go anywhere and hire almost anyone it needs to make its product. Indeed, Tesla is apparently producing cars in Wales, probably because it thinks it has an advantage doing so (perhaps because it's main sales are in Europe where gas is more expensive or perhaps because American unions have threatened to raise costs if they build in America or perhaps because GM wasn't willing to bargain over producing Tesla's at it's own plants).
In short, Mr. Moshiashwili should be admired for his creativity, but the idea should be left where it started -- in the ivory tower.
Posted by: Anonymous Skeptic | March 16, 2009 at 01:30 PM
The analogy is inapt for the reasons you state, but let's not throw out the horses with the bathwater. We have a jobs problems, and we have excess industrial capacity. Also, it is surely noncontroversial to want to encourage start-up enterprises especially in high-tech areas.
Maybe the human capital is too specific for the idea to be tractable, but can't we think about if there are ways to incentivize the automobile manufacturers to accept this sort of work? Maybe the idea should be limited to physical plant and equipment? Would some sort of RFP process solve the identification problems and search costs you mention?
I'm a long way from jumping on board this plan. We're in a huge mess that calls for creative thinking. Let's put all the ideas on the table.
Posted by: Bob Lawless | March 16, 2009 at 02:07 PM
Like Anonymmous skeptic, I applaud your student's creative thinking but engineering wise it isn't very feasible. Each plant at any given time is precisely engineered to make a specific type of car - a SUV or a compact, for instance. To change to another type of car require shutting down production at that plant and reengineering the entire assembly line, among other things. For that reason, it is not feasible to say "now it's Tesla's turn".
Posted by: mark t | March 20, 2009 at 04:52 PM