"The extra hours make it better"

Surely, though, the extra passes improve the work a bit. Even if fear is driving them, the result is a better document. That seems obvious, and it's partly true, so it's worth looking at how much better and at what price.

Joachim Stoeber and Michael Eysenck tested something close to this. They gave 96 students at a British university a proofreading task. The students had to find spelling, grammar and formatting errors in a piece of scientific text, and they could take as long as they liked. The researchers measured each student's perfectionism beforehand, and they timed the task.

Students with high perfectionistic standards weren't found to catch more of the real errors. They did flag more things that weren't errors at all. The researchers call these false alarms. When they divided each student's accuracy by the time taken, high standards went with lower efficiency. Stoeber and Eysenck put it plainly. People with perfectionistic standards, they wrote, tend to find fault even when everything is all right.

That's one study of one task, with students, and it shows a correlation and not a cause. I wouldn't hang a whole argument on it. But it matches what the circles in chapter 1 suggest. After a certain point, a careful person looking for faults starts to find faults that aren't there. "Substantial" wasn't wrong. Neither was "significant". A searching eye with nothing left to find will settle on something anyway.

The more general point is about how improvement is distributed, and this part is my own reasoning. Most of the quality of a piece of work arrives early. The first draft takes it from nothing to rough. The first proper revision takes it from rough to good. After that each pass adds less than the one before. By the fifth pass you're trading an hour for a gain nobody but you could detect. The hours are all the same length, and what they buy keeps shrinking.

And the gain has to be seen by someone to count. The purpose of your summary was to inform six people before a meeting. For that purpose, the version you had on Sunday and the version you had on Wednesday were the same document. No reader would have acted differently on one than on the other. The only place the difference existed was in your own sense of the piece.

You might think this still adds up over a career, and that the person who polishes everything ends up ahead. The evidence doesn't support that. Dana Harari and colleagues pooled the research on perfectionism at work. Perfectionism had sizeable and consistent links with several things that matter in a workplace. Its overall relationship with job performance, though, was what they called equivocal. On average, across the studies, perfectionists weren't measurably better at their jobs. I've read only the summary of that paper, so I can't tell you more about which jobs and which measures. But if the extra hours reliably made the work better, it's the first place you'd expect to see it.

There's a simple reason why it wouldn't show. Time is the one thing at work you can't get more of. The eight extra hours on the summary came out of something, perhaps the supplier calls you didn't make or the thinking you didn't do about next quarter. It may only have been sleep, and sleep pays for Thursday's work. Polishing buys a small improvement in one thing at the cost of a larger loss somewhere less visible.

None of this means revision is a waste. The first revision is often where the real quality comes from, and I'd never talk you out of it. The claim is narrower. There's a line, and past it the hours stop buying anything that anyone will receive.