Yesterday I asked whether we might increase the rate at which people RSVP, thereby improving the planning of parties.
Bob McGrew of CardinalCollective.com suggests a system of raffles. Give out tickets to people who RSVP yes early on and at the actual event choose a winner. Ridicule the winner and do not grant the prize if it turns out he did not show up for the party. Arianto Patunru proposes a related idea. You receive a ticket in the mail, which you activate by calling in a positive RSVP. You then need the ticket to get in the front door. The tickets are read by a machine, so you can’t make personal pleas to be forgiven. And of course you can’t take cell phone calls from your stranded non-RSVPed friends at the entrance.
Bob McGrew also suggests allowing people to respond with a probability of acceptance, rather than a simple yes or no. (It is an interesting question whether people will undershoot or overshoot with their expressed probabilities, I would predict undershooting so the host feels good when the guest actually comes.) Allowing more information might seem like a no brainer but it runs risks as well. People who would otherwise give a definite answer might instead defer their decision and send you a number like 0.84681. And what do you do with all the numbers you get? Estimate a probability density function?
How about the straightforward approach? The old-fashioned Will Baude suggests a greater reliance on social conventions.
My best idea? Have fewer parties. So many social affairs are about signalling in the first place, and we all know that market economies overinvest in costly signals. Stay home and read your favorite blogs.