Organisations are full of forms that nobody defends and nobody removes. A field on an application form asks for information that no decision turns on. A report is circulated to a committee that has not read it in years. A second signature is required on a document the second signatory has no means of verifying. Asked why, the individuals who complete the field and collect the signature say that it has always been done, which is true and is not an explanation.
The explanation is ordinarily available if one examines when the form arrived. Almost every such requirement can be traced to a dispute -- between two departments, between an organisation and an auditor, between a predecessor and a board that no longer exists -- and the requirement is the settlement of it. The second signature was the price of a concession made somewhere else. The report to the committee was what persuaded the committee to relinquish a power. The field on the form was introduced after a case in which its absence was blamed for something, by people who wanted it on the record that they had asked.
Seen in that fashion, the persistence is not inertia. A settlement is a piece of infrastructure: it precludes a question from being reopened, and it continues to do so long after everyone who was party to it has departed, because what holds it in place is not anybody's memory but the fact that removing it would require somebody to declare that the matter is now to be decided differently. Nobody possesses the authority to say that on their own, and the cost of assembling the authority exceeds the cost of the field.
This has a practical consequence for anyone attempting to simplify a process. The customary method is to ask what each step is for, eliminate the ones with no current answer, and be surprised at the resistance. The resistance is not irrational and it is not nostalgia. It comes from people who cannot articulate the original dispute but who correctly sense that the step is load-bearing in some way they cannot name, and who have seen what happens when a settlement is withdrawn without a replacement: the dispute returns, in a new form, to people who do not know it is a dispute they are having.
The better method is archaeological. Find the settlement, establish whether the dispute it settled still exists, and if it does not, say so explicitly and record that it does not, so that the removal is itself a settlement rather than a gap. Processes simplified in this fashion tend to remain simplified, which is the test, and processes simplified the other way tend to regenerate the same requirements under different designations within a few years.
