I like reading articles about how we study voting patterns and how we determine what does and doesn't improve turnout. To be blunt, that's mostly because I think the generally poor turnout in the US is an urgent problem facing our democracy, and because I like having new ways to talk about it and new evidence to point to.
I've never been that happy with the simplest expedient of urging people to vote and shaming them when they don't. It certainly can have an effect, but I think we do better by ourselves and our friends when we look at ways to play up the impact voting has or how to make it easier to do without compromising the accuracy of our vote.
This article is a case in point. It shows how a simple comparison between areas in the US can tell us a lot about this problem, it gives us more information about what sort of benefits are good and which aren't worth the trouble, and it offers a simple solution. That does also suggest that we should be willing to make seemingly large changes in our system; I'm sure some people wouldn't consider changing how and when we submit our votes simple or minor.
That article also does call out one point I think is important: I don't actually believe that fraud is currently enough of a problem to justify measures that will severely depress turnout. Our system can more easily survive fraud at current levels than it can survive even more people not voting because they don't feel they have the time or money to secure everything they need.