Although codetermination is a more frequent scheme of corporate governance among European countries than the mainstream one, founded on agency theory and shareholder value, there is a puzzling dearth of theoretical justifications for BLER (Board Level Employee Representation) and Work Councils, contrasting with the flood of models backing the standard view.

The goal of the paper is to offer an analytical proof of the superiority of codetermination (to be defined more precisely).

The heart of the argument is the dual nature of the enterprise, neglected as much by legal scholars as by economists : the company (UK) or the corporation (USA), i.e. the legal person, whose only members are the shareholders, and the economic organization, commonly called "firm", with its indefinite list of stake-holders, but with an essential contribution made by the workers. Economics only knows the firm, business law only the legal person, whereas the gap between both is the source of a structural coordination problem, between capital and labor.

I show in a 1st part that the standard model of corporate governance where shareholders monopolized the choice of the members of the board is an inefficient solution to the problem of double contractual incompleteness for capital and labor. A fair split of the board would be the efficient one, but its implementation in the real world seem so improbable that the existence of a dozen of european countries with 1/3 BLER calls for a new look about the true nature of the firm.

In order to tackle that riddle, I show in a 2nd part, that the game between capital and labor should be sophisticated through the introduction of a third "agent" or "party", i.e. the management, in charge of the question of coordination between capital and labor within the firm. Work organization becomes a new essential variable (which explains the extension of codetermination towards the integration of works councils) and the whole problem of the firm becomes a dynamic one of organizational learning. Codetermination appears now as a procedural device of creating a common future between capital and labor, guaranteed by a double hostage, since each party directly or indirectly accepts the institutional presence of the other, for discussing strategy and shop-floor organization.

I conclude by considering seven objections.