This past semester I taught “Money and Politics,” an advanced seminar for Government majors concentrating in American politics.  Not only is the class a joy to teach–with some of the best students a professor could ask for–but it also represents a real chance to dig deep into the campaign finance issues of the day.  Aaron Schock was a regular topic of conversation, and a regular source of laughter, this past semester.  No, not for his abs, but for the drip-drip of allegations concerning his campaign finance.  (See here, for example.)  The media did yeoman’s work (and why not, this was a juicy story destined to go viral) in looking at campaign finance records, but the case also makes clear the challenge of using raw disclosure data to really “see” what’s going on.

I downloaded the FEC’s bulk data on congressional candidate expenditures for a different project and coded expenditures into larger categories (e.g., fund-raising, polling, ads, etc).  Expenditures from Schock’s campaign account are attached here, and I’ve highlighted in yellow the expenditures that received some media attention this year.  A few problems in scouring these reports for Schock’s malfeasance are clear:

1. Only some of the reported expenditures were from his campaign account.  Some were from his leadership PAC and some were charged to the government as an expense incurred as a congressman.  So not everything at issue is in the linked file.

2. The reports seem harmless enough.  The top campaign expense is to “Green Chevrolet” for a “transportation expense.”  In this case it was a $73,000 truck, which everyone knows is a must for good constituent service.  But would a regular voter who looked at these reports know something was up here?  Probably not.  He also kept a photographer on salary to help preserve for posterity his adorable adventures, but that also seems harmless in these data.  Those expenses are reported as “salary.”

3. Collapsing expenditures into larger categories is useful, but not for uncovering legal infractions.  According to the FEC’s own codes, Schock devoted 20% of his expenditures to fund-raising.  The mean for House and Senate candidates in 2012 (where I have a complete data set of expenditures coded into larger categories) was 6%.  So Schock was clearly an outlier.  But is that bad?

The case, I think, makes clear that disclosure for disclosure’s sake is only part of the solution.  Bruce Cain has a wonderful new book that asks some tough questions about the kind of populist reforms that recommend more and more and more disclosure, under the notion that data dumps can only serve to enhance transparency and foster better government.  That’s ambitious, though, as a lot can be hidden in plain site.  We need vigorous “linkage institutions” to do the digging for us and connect the dots in compelling ways.  That happened this time, but one wonders….what’s being missed in a world where the media are trusted less and less and polarization makes us suspicious of data and science and expertise?