I don't see the ebb and flow of participation as having a major impact on the cost of delivering an election. We always have to assume that there will be a large number of electors, much more than actually do turn out. For example, the estimate of the number of ballots, the number of electors at a polling site on a particular poll, the number of workers and sites, are all done taking into account to a certain degree an inflated participation because we can't assume people will not show up to vote.
I don't think this will have a significant impact on the financial aspects of preparing and delivering an election.