Correlation doesn't necessarily prove causation. As a temporarily retired poll judge, that low turnout bothers me. The reduction of 2 percentage points doesn't bother me as much as that can be accounted for in terms of swings within a 4 year cycle. Presidential general elections are usually your peaks for turnout with valleys between them and occasional bumps if there is a controversial local issue on the ballot (city charter amendment, trash hauler contract referendum, elected official recall). A reduction in polling places is a one time event but I would be far more interested in seeing it within the focus of that quadrennial cycle.

The other bit of data you'd really want to look at would be the voter registration file. Is there a lot of churn in that file with people coming and going? That can help account for the low turnout. Other demographic details could be in play, too. Not every jurisdiction is as easy-going as Ashtabula County in making their entire voter registration file available as a CSV download: http://www.voterfind.com/ashtabulaoh/vtrreport.aspx

Again, probe a little deeper.