generated sunk costs and networks of political interests that diminish the prospects for radical reform. The possibility of policy preemption suggests that an important source of variation among multitiered systems arises from the timing of interventions by constituent members and central authorities (Pierson and Leibfried 1995: 22).
Though written with European integration in mind, these words apply equally to federal systems like the United States. Policy makers in the U.S. knew that state-level control of workers’ compensation was deeply flawed but chose not to deal with those flaws directly. Proposing any fundamental change to workers’ compensation might jeopardize other valued policy objectives, such as the creation of disability insurance or new national standards for workplace safety. Expecting a bitter fight, national officials chose to leave workers’ compensation alone.
In short, workers’ compensation nicely illustrates several propositions about path dependence in policy making (Pierson 2000). Did timing and sequence matter? Absolutely. The early introduction and rapid diffusion of workers’ compensation laws in the early 20th century were significant, particularly when combined with the slow spread or unavailability of many other forms of social protection. Did seemingly small decisions have large consequences? I argue that one specific feature of early state laws, the reliance on public or private entities to underwrite insurance policies for employers, was indeed portentous. Were policy choices made in the past increasingly hard to reverse? Yes. Did the resulting policy fail to address the underlying problem as well as some other alternative? Almost assuredly yes.
The core of my argument is not new. Years ago the historian Edward Berkowitz, one of the few to study the politics of workers’ compensation in any detail, summarized the program’s history concisely and, I think, accurately:
The fact that lawyers, insurance companies, trade unions, and state industrial commissions all acquired an interest in workers’ compensation has made reform of the program exceedingly difficult .... the program’s basic structure, a product of the Progressive Era, remains unchanged .... Workers’ compensation, then, must be explained in terms of the historical circumstances surrounding its origins. Modern policy problems can be traced directly from the program’s origins in the Progressive Era (Berkowitz 1987: 15-16).
Missing from Berkowitz’s work, however, is a sustained and compelling defense of this judgment. He is far more interested in disability insurance and vocational rehabilitation than in workers’ compensation. My aim in this paper is essentially to prove that Berkowitz was right.