Last December, those looking for signs that Trump associates had colluded with the Russians to influence the presidential race got a new piece of information to support their case. In a court filing that was improperly redacted, attorneys for Paul Manafort inadvertently revealed that their client had allegedly shared polling data pertaining to the 2016 race with Konstantin Kilimnik, a political operative with whom Manafort has previously worked, who purportedly has ties to Russian intelligence.
In this week’s sentencing memo from Special Counsel Robert Mueller, we got another tiny additional detail on the polling that was shared: a page count. The Daily Beast reports that the polling data, labeled “Exhibit 233” in court documents, appears to be 75 pages in length. When the initial news broke that Manafort may have shared polling information, there was a wide range of things that could mean; a detail as small as page length can slightly narrow the possibilities of what exactly Manafort allegedly shared, and for what purpose that data could have been used.
For those outside the polling world, 75 pages may sound like a lot of information, but it depends entirely on the form the information takes across those pages. Typically, a pollster produces a handful of items for their client that could be circulated to others and characterized as “polling data,” ranging from polling data in its rawest form all the way up to presentations and strategy memos built around that data. While it would be unusual for a campaign strategist to directly share internal campaign polling with someone unaffiliated with the campaign, some forms of sharing would be more unusual than others. Given this, it is helpful to consider which of these polling data formats is most likely to result in a 75-page Exhibit 233.
The first things a pollster typically sends their client are “toplines” and “crosstabs.” Toplines are the overall results for each question of the survey. (You can see an example here.) While it would be very unorthodox for a campaign manager to shoot such a document to someone like Kilimnik, it isn’t hard to imagine someone like Manafort dropping them as an attachment in an email to someone he wants to impress. However, even for a very lengthy survey, toplines are unlikely to stretch more than 20 or so pages. For there to have been 75 pages worth of polling information, we’re either talking about toplines from multiple surveys, or we’re talking about toplines in addition to something else.
Crosstabs, on the other hand, are usually much longer. In fact, 75 pages would be on the quite slim side for a set of crosstabs, depending on how they are formatted. Crosstabs allow a reader to look at every question asked in the survey, and its result among any number of subgroups of survey respondents. (You can see an example here.) This usually means you have multiple pages for each question asked, and each page has dozens upon dozens of data points on it. You can think of crosstabs like a phone book: very thick, dense, and full of information, only some of which you will ever actually end up using. A set of crosstabs that is only 75 pages would mean the survey was fairly short and the data was not being sliced very much, making it possible but unlikely that Exhibit 233 is just a skinny crosstab book.
Finally, there’s the analysis: the memos and slide decks that can accompany the results. These can vary widely in length, from a two-page memo looking at the key takeaways from the survey all the way to an extensive slide deck analyzing the results and offering strategic recommendations. Could an extremely thorough slide deck reach 75 slides? Absolutely. Most damning for Manafort, if “Exhibit 233” is a slide deck of that length, it almost certainly includes information of strategic value on topics like messaging and voter segments. While the sharing of toplines or a slim crosstab book would be of limited utility to the recipient as a roadmap for election meddling, an extensive report deck would have a lot of information about a campaign’s strategy.
So long as it remains redacted, we don’t know what Manafort allegedly shared with Kilimnik, which makes it hard to know how valuable it might have been to whoever else might have obtained it. Whether that data is in topline, crosstab, or report-deck form makes a big difference in just how strategically valuable those 75 pages might have been to their recipient.
