Via Michael Makowsky, there is a new and extremely useful paper by Chetty, Saez, and Sandor in “slides” form (pdf). Their conclusions include these:
1. Short deadlines for referees are extremely effective at increasing speed.
2. Cash incentives can generate significant improvements with salient reminders right before a deadline.
3. Even light social incentives, such as direct prods from an editor, can bring significant benefits.
More broadly, at least in this context cash incentives work, they do not displace social incentives, and attention really matters as do “nudges.”
I really like the public shaming treatment of listing referee names with turnaround times (and the feigned disgust with public shaming [and low incentives ... however, it is Elsevier so they could afford more] as an excuse to decline a review!).
Another tip might be to expand the pool of referees. It seems like everyone has a constant stack of 3-4 papers on their desk. My impression is that there are a boatload of competent referees that simply refuse to referee papers. Since refereeing is a voluntary contribution to a public good, I have no idea how to increase those contributions.
I also like the $100 Amazon Gift Card incentive. That said, the accounting policy at Appalachian State University is that no survey or incentives can be greater or equal to $100. Because all hell breaks lose if someone gets a $100 incentive (e.g., it could be one of the signs of the apocolypse), know what I mean?
But what I really want to know is, is there a written paper? I've presented several "papers" in "slides" form that never made it to being a real (written) "paper."