There's a bit of a discussion in the comments to the next post down about why GM failed. Since that's what I'm studying (except not GM, I'm just studying generally why some firms outperform other firms), I probably ought to have an opinion. And I do.
It was management's fault. It's always management's fault. It's in the job description. But that's not a particularly satisfying or interesting answer.
I'm not sure why GM died, but there's one explanation (or rather one set of explanations) that's worth looking at. I think that it's fair to say that most strategic management scholars (and I) think that strategy matters. That organizations can effect their own performance by choosing strategies that fit their environment. The sub-field I'm working in, strategy process, is the study of how organizations monitor their environment, move information to the places in the organization best able to act upon it, and then develop strategies to fit the current environment. This side of strategy assumes strategic choice.
But there is a another side of strategy called, not entirely appropriately, determinism. These theories basically assume that strategy doesn't make a difference; that individual organizations don't really have much control over their own performance. Michael Porter, who some of you have probably heard of, started strategy by more or less teaching that what industry a business is in makes all the difference -- all companies can do to help performance is choose the right industry and then work with their competitors to shape the industry by, for example, getting the government to restrict entry. Other scholars think that organizations are subject to the laws of population ecology; or that corporate routines are like genes, immutable in the organization but mutable as new organizations are created. Other scholars think that young organizations can be entrepreneurial, but then bounded rationality, risk and loss aversion, satisficing, etc., embed the existing strategy which cannot thereafter by changed. My off-the-cuff reaction is that this last description fits GM.
Miles and Snow (1978) -- a great book I recommend to lay business people and Harry -- divides up corporate functions between the entrepreneurial, the engineering and the administrative. Different companies emphasize different functions, and the result is four types of companies: prospectors, analyzers, defenders and reactors. Any of the first three can be fine, depending on the environment. Being a reactor is fatal. GM was founded to concentrate on the administrative function. Toyota and Honda focus on the engineering function, with some attention paid to the entrepreneurial function. GM never could shake its focus on the administrative function; they always thought that they could manage themselves out of the predicament. Over time, they slid from analyzer to defender to reactor, and then they died. This makes it sound as if GM could have saved itself if it had become entrepreneurial, or if its unions had allowed it some engineering slack. But GM was what it was. There's no sense telling the lobster that it could save itself from the pot if it would only sprout wings.
Of course, that suggests that the government bailout was exactly the wrong thing to do.