diffusion of state workers’ compensation laws; the heavy reliance on private employers and insurers to deliver benefits; and mounting evidence that state laws had serious imperfections that the national government would find costly to remedy. Once the program had been excluded from the Social Security Act, it was so embedded politically and institutionally at the state level that future reformers found it almost impossible to imagine a successful campaign to national or even federalize the program. In a number of key respects, such as the mix of public and private insurers and the wide variation in benefits by state, workers’ compensation has changed remarkably little since the 1930s.
It is important to remember that workers’ compensation did not remain the province of the states simply because of a dense network of stakeholders. National officials never contemplated comprehensive reform to workers’ compensation as a single, separate issue. It was always bundled with other issues such as the Social Security Act, disability insurance, and new occupational health and safety regulations. The history of workers’ compensation cannot be understood apart from the history of these other initiatives, for policy makers repeatedly chose to leave workers’ compensation alone in order to improve the odds of winning these other battles. Charting a new path for the public sector was hard enough without simultaneously trying to redirect an old path.
There was nothing inevitable about this history. Widely shared perceptions of a “crisis” in workers’ compensation could have given rise to a strong push for national involvement. In that case, the cost of switching paths might have seemed less daunting. Instead, the costs were repeatedly magnified by the potential of undermining some related and valued expansion of the public sector. Thus, this case suggests that arguments about path dependence in policy making must be sensitive to the ways in which actors’ preferences and goals can be shaped by the trajectory of two or more policy paths.32
The implications for our understanding of policy preemption are more ambiguous. Pierson argues persuasively that one of the unappreciated trends in U.S. social policy making has been the gradual nationalization of income transfers since the New Deal. Some of this transformation has been due to the creation and expansion of new national programs like Medicare and the Earned Income Tax Credit, and some to the rapid growth of older national programs like Social Security. The more intriguing part of his argument concerns patterns of success (Food Stamps, SSI) and failure (unemployment insurance, AFDC) at nationalizing social programs. Pierson argues that these battles were waged less over jurisdictional questions than over substantive changes in policy content. In other words, the issue was not state versus national control, since state officials had little to gain politically by retaining control of income transfers to the poor.33 The issue was whether coverage would be expanded, benefits made more generous, and both made more uniform. “Opposition stemmed far less from the desire of state officials to amass and retain bureaucratic power than from the demands of many interests to prevent a nationalization of benefits that would hinder economic strategies based on low wages and stingy social