An August 21 article in The New York Times about the Romney campaign’s fundraising reflects both of these problems:

But the biggest outside spenders in the advertising battle are tax-exempt groups that do not have to disclose their donors. In the last 30 days, Americans for Prosperity, founded by the billionaire industrial executives Charles and David Koch, has spent at least $14 million on ads against Mr. Obama, F.E.C. records show. Crossroads Grassroots Policy Strategies, a nondisclosing affiliate of American Crossroads, spent more than $16 million, according to the [Kantar Media] Campaign Media Analysis Group, which tracks ad spending.
In fact, Kantar and FEC data show widely disparate totals when measuring (again, ostensibly) the same thing for these groups: For example, Kantar currently shows Americans for Prosperity spending $7.5 million more on TV ads than the FEC does.

To figure out why Kantar and the FEC’s statistics vary so dramatically, CJR spoke with Kantar Media, The Washington Post, and the government transparency groups Sunlight Foundation and Center for Responsive Politics. Below are the likely reasons we’ve identified as to why spending totals in the TV ad wars vary so widely between these two sources:

Future Ad Buys. There is a substantial category of spending that would be covered by FEC data but not by Kantar: TV ads that have been purchased but have not yet aired. Outside groups (and campaigns) pay for ads at the time they reserve them, and must report to the FEC at the time of the expenditure. By contrast, Kantar tracks campaign ads as they run, and their spending totals reflect ads that have already hit the airwaves.

Gaps in Kantar coverage. Kantar tracks campaign ads that run on national networks, local network affiliates, and national cable. But it does not include ads that run on local cable stations. Elizabeth Wilner, the vice president of Kantar’s Campaign Media Analysis Group, estimated based on her own analysis in an August article in Ad Age that overall spending in the presidential race would hit $200 million on local cable—a chunk that Kantar’s tally misses entirely.

Cost underestimates by Kantar. Kantar measures the volume of TV ads that run, and then, to estimate their cost, uses a formula based on the average ad rates for programs in which they aired. But many of Kantar’s estimates do not account for a crush of campaign ads—such as in the days before Republican presidential primaries—driving up the cost of each spot to significantly higher than the average. It does presume higher rates in the fall due to political spending.

Disclosure loophole for “issue ads”. The FEC also has significant gaps in its data. The Commission allowed nonprofit groups such as Americans for Prosperity to avoid reporting substantial spending through a loophole that exempted so-called “issue ads,” which do not specifically tell voters to support or oppose a candidate, from disclosure unless they run within 30 days of a primary or 60 days of a general election. (Here’s an example of one such ad by Americans for Prosperity, entitled ”Wasteful Spending.” It concludes: “President Obama wasted $34 billion on risky investments. The result: failure.”)

Commissions for ad placement firms. Campaigns and outside groups do not buy ads directly from broadcasters. Instead, they make payments to ad placement firms who act as middlemen in purchasing TV time. The firms charge a commission that can be as high as 10 percent of the ad buy, experts told CJR. The cost of these commissions would be captured in FEC data, which reflects overall expenditures, and not in Kantar data, which reflects the costs of the ads themselves.

Imprecise disclosures to the FEC. FEC spending disclosures are broken down by purpose, such as “Advertising-TV Placement,” “Web Ads” or “Voter Contact Phones.” But these designations are not uniform between groups, and sometimes are so imprecise that they do not clearly describe how money is being spent. For example, as we noted in our graphic, the pro-Romney super PAC Restore Our Future has spent $75.1 million on “Media Buys.” This is a separate category from Internet ads and print ads, but we cannot be fully certain that it is all dedicated to television and none of it goes to radio.

Genuine misconduct. The litany of discrepancies and imprecisions above opens the door for real misconduct. If the reason the FEC was showing higher TV spending than Kantar for Priorities USA was, in fact, that Priorities USA was misreporting the purpose of their expenditures, or if Kantar was actually showing higher TV ad spending for Americans For Prosperity because AFP had failed to report relevant material to the FEC, we would be none the wiser.

Sasha Chavkin covers political money and influence for CJR's United States Project, our politics and policy desk. He has written for ProPublica, the Center for Public Integrity, and The New York World. Follow him on Twitter @sashachavkin.