The Science of "Muddling Through" Charles E. Lindblom Public ...

88 PUBLIC ADMINISTRATION REVIEW overlook excellent policies for no other reasonthan that they are not suggested by thechain of successive policy steps leading up tothe present. Hence, it ought to be said thatunder this method, as well as under some ofthe most sophisticated variants of the rootmethod-operations research, for examplepolicieswill continue to be as foolish as theyare wise.Why then bother to describe the method inall the above detail? Because it is in fact acommon method of policy formulation, andis, for complex problems, the principal relianceof administrators as well as of otherpolicy analysts.9 And because it will be superiorto any other decision-making methodavailable for complex problems in many circumstances,certainly superior to a futile attemptat superhuman comprehensiveness.The reaction of the public administrator tothe exposition of method doubtless will beless a discovery of a new method than a betteracquaintance with an old. But by becomingmore conscious of their practice of thismethod, administrators might practice it withmore skill and know when to extend or constrictits use. (That they sometimes practice iteffectively and sometimes not may explain theextremes of opinion on "muddling through,"which is both praised as a highly sophisticatedform of problem-solving and denounced as nomethod at all. For I suspect that in so far asthere is a system in what is known as "muddlingthrough," this method is it.)One of the noteworthy incidental conse-Elsewhere I have explored this same method ofpolicy formulation as practiced by academic analystsof policy ("Policy Analysis," 48 American EconomicReview 298 Dune, 19581). Although it has been herepresented as a method for public administrators, it isno less necessary to analysts more removed from immediatepolicy questions, despite their tendencies todescribe their own analytical efforts as though theywere the rational-comprehensive method with an especiallyheavy use of theory. Similarly, this same methodis inevitably resorted to in personal problem-solving,where means and ends are sometimes impossible toseparate, where aspirations or objectives undergo constantdevelopment, and where drastic simplification ofthe complexity of the real world is urgent if problemsare to be solved in the time that can be given to them.To an economist accustomed to dealing with the marginalor incremental concept in market processes, thecentral idea in the method is that both evaluation andempirical analysis are incremental. Accordingly I havereferred to the method elsewhere as "the incrementalmethod."quences of clarification of the method is thelight it throws on the suspicion an administratorsometimes entertains that a consultantor adviser is not speaking relevantly and responsiblywhen in fact by all ordinary objectiveevidence he is. The trouble lies in the factthat most of us approach policy problemswithin a framework given by our view of achain of successive policy choices made up tothe present. One's thinking about appropriatepolicies with respect, say, to urban traffic controlis greatly influenced by one's knowledgeof the incremental steps taken up to the present.An administrator enjoys an intimateknowledge of his past sequences that "outsiders"do not share, and his thinking andthat of the "outsider" will consequently bedifferent in ways that may puzzle both. Bothmay appear to be talking intelligently, yeteach may find the other unsatisfactory. Therelevance of the policy chain of succession iseven more clear when an American tries todiscuss, say, antitrust policy with a Swiss, forthe chains of policy in the two countries arestrikingly different and the two individualsconsequently have organized their knowledgein quite different ways.If this phenomenon is a barrier to communication,an understanding of it promises anenrichment of intellectual interaction in policyformulation. Once the source of differenceis understood, it will sometimes be stimulatingfor an administrator to seek out a policyanalyst whose recent experience is with a policychain different from his own.This raises again a question only brieflydiscussed above on the merits of like-mindednessamong government administrators. Whilemuch of organization theory argues the virtuesof common values and agreed organizationalobjectives, for complex problems inwhich the root method is inapplicable, agencieswill want among their own personnel twotypes of diversification: administrators whosethinking is organized by reference to policychains other than those familiar to most members.of the organization and, even more commonly,administrators whose professional orpersonal values or interests create diversity ofview (perhaps coming from different specialties,social classes, geographical areas) so that,even within a single agency, decision-makingcan be fragmented and parts of the agencycan serve as watchdogs for other parts.