I did not mean to suggest that a university should be organized in the same way as a conventional business firm. For that matter, there is no single way in which business firms are or should be organized; optimal organization varies with the nature of the business. In a software company, or any other company whose success depends on the imagination and originality of its "knowledge workers," hierarchical control will be loose and employees will have more scope and say. And so in a research university like Harvard we expect the faculty, the principal employees--and knowlege workers all, to have a large say in governance. But their say may well be too large. (Teachers are much more influential in public schools than in private schools--is that a good thing?) One reason U.S. universities are better than European universities is that the latter are even closer to the workers' cooperative model of organization than our universities are--often the president of a European university will be elected by faculty, students, and staff. Students are birds of passage and immature; most staff knows little about the institution; and university faculties--to be blunt--contain not only a fair proportion of social misfits (for academia places little value on interpersonal skills), but have become so specialized that most academics know and care only about their own, often very narrow, field of research.
Larry Summers can be seen as attempting an incremental shift in university structure toward the business model. Compare a university department with the corresponding division of a business firm. The department is likely to operate as a self-perpetuating oligarchy, with the chairman appointed by and dependent on the faculty of the department and owing all fealty to the department rather than to the university. Rarely will a department chairman aspire to a leadership role in university administration, and so he will have no incentive to bring a university-wide perspective to his job. In contrast, the corporate division head will be looking to rise in the corporate hierarchy, and this will require him to manage his division with due regard for the division's role in the corporation as a whole.
There is room for experimentation with tightening university control over departments, as by empowering the university's president to appoint and remove department chairmen, and empowering department chairmen to make occasional faculty appointments without a departmental consensus. The result might be more responsible, coherent management.
It is plain that with all their strengths, American universities have plenty of problems. Neglect of teaching by academic stars is one. The reign of political correctness, which makes a mockery of the academic commitment to wide-open debate, is another; a related point is the political imbalance of university faculties. The continued advance of specialization, which threatens to destroy any general intellectual culture and further estrange the universities from the society as a whole, is a third problem. Universities need better management to solve these problems, and better management may require changes in organizational structure.