Paradoxes of Welfare-State Conservatism
Teles, Steven M., The Public Interest
AT times an unbridgeable gap seems to separate the policy sciences from the practical world of politics. Politicians complain that policy analysis is not "useful" to them, and policy analysts lament what they see as politicians' ignorance of even the most rudimentary analytical principles. This gap is not simply a matter of willful obscurity on the one side or policy illiteracy on the other. Rather, it reflects a conscious choice of policy experts to eschew political and institutional analysis in favor of a more pure science of policy.
The costs of such a choice can he seen most clearly when analyzing America's largest social-welfare program--Social Security. Critics of social insurance make arguments that are by now astoundingly common; their techniques of analysis are almost always dependent upon an accepted, but morally and politically suspect, set of assumptions. Moreover, these assumptions are radically utopian, and thus a dangerous foundation for reasoning about social institutions like the modern welfare state which are deeply embedded in social life. I will lay out, in the context of a discussion of pensions, a way of approaching welfare-state questions that differs from modern libertarian arguments and also from the "progressive" arguments that have dominated welfare-statist approaches up to the present.
My argument is that one of the outcomes of a public policy is the politics it creates. Policies are more than mechanisms for the solving of problems: They are also contexts in which future policy battles will be fought. Put another way, policies do not simply operate within a context--after they are implemented, they are the context. If this is true, it is impossible to design policies rationally without taking politics into account. If the outcomes of public policies are both their effects on clients and on the character of public debate and institutional development; then the artificial academic division of labor between "policy analysis" and "policy process" must be collapsed. Introducing political considerations necessarily adds an element of indeterminacy to the process of policy analysis, one which goes band in hand with a temperamental conservatism. But one might ask whether such a thing as conservative policy analysis exists, and if it does, whether it could tell us anything useful about the entitlem ent reform debate in the United States and elsewhere. Understanding conservatism properly, the answer to both questions is yes.
Politics and policy analysis
What is the relationship between politics and policy analysis? It is obvious that policy analysts are not "philosopher kings," handing down perfect policies to a waiting world. In fact, politicians regularly ignore the recommendations of policy analysts, making decisions that the latter consider "irrational," or "not informed by the evidence," or "not the best practice."
One solution would be simply to ignore the world of politics, to hide behind the mask of science. This approach views policy analysis as akin to basic science and the political world as akin to engineering. Should we change our recommendations just because the engineers keep screwing up? Isn't it our job to discover the truth, and if the world ignores it, so much the worse for the world?
This would, of course, be one solution, and it is tar less rare than my (admittedly cartoonish) description might suggest. But policy analysts are not like basic scientists: Our very job is to make recommendations, and the very act of recommending includes some sense that those being advised, the politicians, must be taken into account. But how?
A second approach would view the political world as a series of obstacles to the "one best way." That is, policy analysts must start out by determining "the truth," but they would then take into account the need to have their proposal passed by real legislators and administered by real bureaucrats. …