Since everyone is familiar with scenarios where you shouldn’t procrastinate, let me start with two examples where I endorsed procrastinating.
I was teaching an art class for the first time, which ran across three days. I had plenty of materials gathered but had yet to make a final selection on which topics to teach, which examples to use, and what specific rules I should give for exercises I had in mind. The class took place in an enclosing week-long event, so working on my laptop directly traded off against socializing in the venue. Class was started the next morning, two hours after my wake-up time, and I still struggled to finalize my material. After wasting an evening this way, I decided that I would leave it for the 90 minutes before class started.
This was a great choice. The indecision and dithering and hopping from section to section fell away, and in slightly under an hour I had a day’s worth of material. I repeated this the subsequent two days.
(It is important in this case that I’d gathered my materials beforehand and could imagine the task in detail – that is, the pre-work was done. I was ready to launch into it the moment I had to, and I knew roughly how long it would take.)
My second example is reviews. For some years I was a big fan of David Allen’s GTD task system, and I believed Allen that regular review of your priorities and todos was essential to making it work. I still think it’s a good idea if you can do it, but I could not, because my performance on the task was so abysmal when I wasn’t in the mood.
A life review is not like the dishes, which end up clean even if I wash them while not in the mood. A review I do while not in the mood will drain precious energy and be unhelpful to boot.
The obvious principle that determines when you ought to procrastinate is comparative advantage. Is it your present self’s comparative advantage to do it? If yes, obviously do it. If it seems exactly flat, also do it right now. If no, punt it.
Why do it right now if it seems exactly even? Because if your present self does it, you will have had the thought you should do it just once, right now. If your future self does it, you will have had the thought twice. Not just the thought, but the attendant indecision and subtle confusion too. You can increase your subjective lifespan by spending less of your time noticing and then un-noticing tasks that can be done anytime.
Comparative advantage can be internal or external. Internal is usually about whether I’m in the mood. I write this the day after I stayed up until 3am responding to an important email I’d been avoiding for five days. I’d tried to force myself to write it the day I got the email, but produced a very flabby response. When the mood struck me, however, several thousand words flowed out of me.
(I endorse being prompt with email, so I should have done this on purpose, which for me means trying at 1am. I’m less neurotic and more verbal late at night.)
External comparative advantage for me is sometimes about context (where I am, who I’m with, what devices I have) but usually about deadlines, as with the art class lesson plan. It’s reasonable to wait until a deadline is close to start work. But this is much safer if you don’t procrastinate on the pre-work, and know what the work will consist of and what the prerequisites are.
(Procrastinating until a programming project deadline is particularly pernicious because it is difficult to simulate what the work consists of.)
There’s a principle that falls out of this which runs counter to normal ‘organized person’ task triage systems like GTD: if you’re in the mood to do a task you usually don’t want to do, you should err on doing it even if it’s not that important.
Most people would benefit from GTD, but a perfect adherence to GTD is not optimal because it wastes energy.