[euyyn]

I could imagine this:

"Oops, I forgot about presenting my subjects with the consent form and now I've got 100 results collected during a whole year of work.

I definitely don't want to throw all of that work away... But if I contact them and make them sign the form now using pencil, I could erase the date and change it to each subject's day of experiment. Nobody would notice."

Pen raising that barrier. I agree that against an adversary fully committed to fraud it doesn't matter. But that threat model is exceptionally rare, and an IRB can do little about it anyway (e.g. the whole application could be made up, the experiment consisting of something completely different). The realistic threat they fight against is experimenters that don't know better, and think there'd be no harm in their actions.

And this is just me coming up with a possibility in a few minutes. Reality tends to be richer than a single person's imagination.

[yebyen]

Some researcher who paid $40 to 100 people over the course of a year is going to go back and contact all of those same people again, to sign a release form, and they all have to use pencils... and nobody is going to call him out on it and ask why is it important that I sign this form with a pencil?

The fact that I actually agree with you, and think this might work in fact inside of any hospital in the US, says more about the byzantine complexity and scary state of our medical system than anything else.

Because I could totally see 100 people being told they have to sign a thing, while in a hospital, with absolutely no good reason given, ... and all 100 just signing it after some variable x*N amount of huffing, then get on with their day.