Douglas Stevens believes that the balance between corporate autonomy and institutional control is a central issue in the management of Crown corporations. In Corporate Autonomy and Institutional Control he investigates the effects of various organization designs on this balance, taking into account the incentives and strategic behaviour of vote-maximizing politicians, influence-maximizing bureaucrats, and autonomy-maximizing managers.

Stevens examines institutional frameworks for Crown corporations in Alberta, Saskatchewan, and Manitoba between the early 1970s and the mid 1980s, showing how each framework establishes different practices and offers distinct strategic advantages. Organizational approaches in Alberta most closely approximated what the author calls a "self-contained" design, in which corporate actors had the advantage and were most able to achieve their own objectives. In Manitoba, where "vertical information systems" prevailed, central bureaucratic monitoring agents tended, to some extent, to wield influence over the corporations. Saskatchewan practice was akin to a "lateral relations" pattern, with an equilibrium between corporate and bureaucratic goals.

Stevens's comparison of Crown corporation organization designs suggests that, while no one form is inherently more efficient than another, each leads to qualitatively different outcomes. He concludes that the most important issue in problems of organization design is who is winning the Crown corporation "game" -- a finding of considerable interest to all students of government enterprise.