Who the passive voice hides
The grammar is not sloppy. It is doing exactly what it was chosen to do, and it is doing it to a specific person.
Contents
It was noted that the reconciliation had not been performed on a timely basis.
I could parse that. I could not answer any question about it. Who noted it. Who did not perform it. Timely by whose definition, agreed with whom. The sentence contains a finding and no people, and it is not badly written. It has been written carefully by somebody who knew exactly which words to leave out.
Sentences in that register filled the audit reports I assessed at the Canada Revenue Agency, in the first job I had, against the auditing standards and the internal control policy. I had expected the difficulty to be technical. Some of it was. Most of the difficulty, for the first stretch of it, was that I could not tell what the sentences were saying.
The removal is the content
English lets you delete the actor from a sentence and keep the event. Something was found. A control was not operating. An adjustment was made. The action survives and the person evaporates.
In institutional writing this is not laziness and it is not a stylistic tic inherited from a manual. It is a decision, and once you see it as a decision you can read the document backwards.
Every place the actor is missing, ask who would have been named there, and the answer tells you what the document is protecting. Sometimes it is protecting a named individual, and the passive is a kindness that costs the reader information. Sometimes it is protecting the writer, because naming the actor would require them to have established who it was and they did not. Sometimes it is protecting a relationship the writer needs to survive the report, which is the most common one and the least discussed.
And sometimes, genuinely, nobody knows. A process failed across three teams over four months and there is no actor, because there was no actor. That is a real finding and it deserves a sentence that says so, and instead it gets the same grammar as the three cases above, which is how the useful version becomes invisible.
Rewrite two sentences with people in the margin
The fastest read I have on any report is to find the sentences with no people in them and rewrite two of them with people in.
Not out loud. On the page in front of me, in the margin. The reconciliation had not been performed becomes somebody did not perform the reconciliation, and the next question writes itself, and it is a question the report went out of its way not to prompt.
Half the time the rewrite is impossible, because the information genuinely is not there. That is worth knowing too. A finding you cannot attribute is a finding at a lower confidence than the sentence implies, and the grammar is concealing the confidence level along with the actor.
The part where I was on the other side
I wrote the monthly working papers behind the trial balances. I wrote them in exactly this grammar.
Not to hide anything from anyone. I wrote it that way because it was the house style, because every document I had read was written that way, and because writing an amount was adjusted takes less thought than writing I adjusted the amount after speaking to the person who owns the account, who told me the following. The passive is shorter. It is shorter because it carries less, and less is easier.
That is the honest account of most of it. The construction survives mainly because it is cheap, and the concealment is a side effect that a small number of people then learn to use deliberately.
Which is worse in a way, because it means the document that is genuinely hiding something looks exactly like the four hundred documents that are merely tired.
Where the habit stops working
I have made myself unpleasant with this. There is a mode where you demand an actor for everything, and it reads as an accusation even when it is not, and people start writing you defensive prose that is worse than what you were complaining about.
There is also a version of institutional writing where the removal of the actor is correct. A standard should not say I think this is enough. A rule that applies to everyone should be written as though nobody in particular wrote it, because the alternative is a rule that expires when its author leaves.
So the same construction that hides a person is also the one that lets a rule outlive one. I have not found a clean way to tell those two apart in advance, only afterwards, by whether anybody was ever able to argue with it.