The Texas Politics Project

Just how rational are bureaucracies when they make decisions--and what do we even mean by rational? The ideal view of bureaucracies envisions highly organized and thorough organizations that systematically evaluate a wide range of alternatives and then choose the best of these alternatives based on careful analysis. But a more realistic analysis suggests that bureaucracies often make decisions based on a much more limited range of information and analysis. In one of the earliest formulations of this view, bureaucracies make decisions by "muddling through."

Charles Lindblom's classic article "The Science of Muddling Through" (1959) outlined his view that the U.S. executive bureaucracy uses limited policy analysis, bounded rationality, and limited or no theory at all in formulating policy. In some ways, Lindblom and those who developed his interest in streamlined decision making in bureaucracies presaged more recent attention to the ways that individuals make very quick decisions using very little information. (Malcolm Gladwell's 2005 book Blink gathered research from a range of disparate fields to explore how "snap judgments" commonly use very abbreviated information, like facial expressions, to make judgments in the space of a few seconds.)

Lindblom took issue with the so-called "rational-comprehensive" approach that had dominated research and teaching in public administration, and still is often presented as the ideal method for developing public policy. According to this approach policy makers begin addressing a particular policy issue by ranking values and objectives. Next, they identify and comprehensively analyze all alternative solutions, making sure to account for all potential factors. In the third and final step, administrators choose the alternative that is evaluated as the most effective in delivering the highest value in terms of satisfying the objectives identified in the first step.

This approach seems to make perfect sense. But bureaucrats and administrators don't work this way in the real world, according to Lindblom.

First, defining values and objectives is very difficult. There are always trade-offs in public policy. It is difficult to say with certainty, for example, that it is better to spend less on education in order to balance the budget. Or that building more roads is a better way to reduce traffic congestion than raising gasoline taxes. Or vice versa.

Second, separating means from ends (policy recommendations from the objectives of those policies) is impossible. Instead, the policy solution is always bound up with the objectives. The problem of reducing traffic congestion could involve building either highways or mass transportation. But for many interested parties each of these potential "solutions" to the problem of congestion is likely to be a policy goal in its own right.

Third, it is impossible to aggregate the values and objectives of the various constituencies of the executive bureaucracy--citizens, private organizations, legislators, and appointed officials, among others--to determine exactly which preferences are most important. The virtue of a policy is indicated by its ability to achieve broad support, not by some assessment that it is most efficient according to some abstract criteria.

Finally, it is inefficient to identify and analyze every policy option. For all but the most narrow policy choices it takes too much time and too many resources. Administrators are very busy and the volumes of detail on even relatively simple issues would be overly burdensome to analyze.

Instead of comprehensive analysis of every policy option, a much more constrained process of "successive limited comparison" is really how policies are developed, insists Lindblom. According to this "branch" method, administrators usually look only at policies that differ in relatively small degree from the policies currently in effect, thereby reducing the number of alternatives to be investigated while simultaneously narrowing the scope of investigation. In other words, they look at two nearby branches, not the whole tree, roots and all.

Successive limited comparison--or muddling through--is thought to be the primary cause of the tendency toward incrementalism in policy development. Only rarely are dramatically different new policies developed. Instead, administrators in both the public and private sectors tend to build on existing policies, tweaking them here and there in a continuous, evolutionary process.

This sometimes causes frustration on the part of citizens and other interested parties, who feel that the government is sluggish and unresponsive. But, Lindblom thinks that such incremental "muddling through" is a good thing. It is efficient (it analyzes practical options much more quickly than the root method) and in the end it is responsive to the goals of a sufficiently broad set of constituents.

Maybe so. But, couldn't he have given this process a better name?