The immediate impact of the collective voice model on research into the economic consequences of unions is hard to exaggerate. Up to the publication of What Do Unions Do? much of the mainstream economics profession viewed unions as combinations in restraint of trade, as monopolies (almost) pure and simple. Although organization theory had long suggested that positive union impacts on firm performance might result from shock effects (Leibenstein, 1966) — that is, having to pay a union premium shocks management into looking for cost savings elsewhere, in the process eliminating or reducing slack within the organization — neoclassical economists were generally not only leery of shock effects but also prone to emphasize union restrictive practices as the source of X-inefficiency. 1 In this accessible, monopoly view of the world (and abstracting from the costs of the union rule book), unions were viewed as having adverse effects on efficiency by distorting factor prices and usage, redirecting higher quality workers and capital from higher to lower marginal product uses. Furthermore, to the welfare triangle loss(es) had to be added some portion of the transfer effect, as unions engaged the polity to protect their monopoly powers. To be sure, the costs of strikes were no longer uncritically laid at the door of unions, 2 but there were already sufficient distortions associated with the union entity to render this advance of marginal interest only.