Yes, well, that's a good question, and I've not spent a lot of time thinking about what that would look like. But I think if one were to imagine what kind of study would be useful, one example would be to put in place the proposed change--let's say we're looking at advanced polling on the Sunday prior to the election and on the Sunday prior to that--in selected constituencies on a trial basis and actually measure the impact on voter turnout in those constituencies relative to constituencies overall. I mean, you would design a research study around the election event itself and be in a position to get a better estimate of the impact of that change.
Another could be to introduce an initiative like, oh, I don't know, Rock the Vote, for example, which is increasingly widely used by election administration authorities. They sponsor music events that are designed to encourage young people to come to those events and then use those events as opportunities to increase political awareness. You would look at the effectiveness of participation at an event like that in respect of the cost of the event.