• by Joe Mandese  @mp_joemandese, September 6, 2024

If the race for the White House was one of Donald Trump's reality TV shows, he'd be losing by a metric that has historically mattered most to him. No, not just his weird obsession with crowd sizes, but by one of the most important currencies in the ad industry: Nielsen ratings.

It's well-known by now that the Democratic National Convention attracted 15% more Americans than the Republican National Convention, but a special Nielsen analysis conducted for MediaPost reveals a Democratic advantage in an even more vital geographic segment: the seven so-called swing states that are expected to determine the electoral outcome, deciding who actually gets elected: Trump, or Kamala Harris.

"When you look at this data, I don't think 1% or 2% is statistically significant. That's virtually a dead heat," explains Nielsen Senior Vice President-Product Strategy & Thought Leadership Brian Fuhrer, who conducted the analysis (see details at bottom) and says that, to the best of his knowledge, it is the first time Nielsen has ever released swing-state ratings data publicly.

"What leapt out to me," he continues, "was Georgia, and obviously total U.S., but also Pennsylvania and Arizona. I don't think those percentages are nothing."

While the focus is on the seven so-called "battleground states" -- Arizona, Georgia, Michigan, Nevada, North Carolina, Pennsylvania and Wisconsin -- which amount to 93 of the 538 Electoral College votes, some matter more than others.

In other words, if Nielsen ratings are a proxy for voter interest -- if not explicit voting intent -- the national ratings are an indicator of the popular vote, while the states' ratings analysis is a signal for electoral votes.

The three statistically significant states cited by Nielsen's Fuhrer -- Arizona, Georgia and Pennsylvania -- will contribute 46 of the 270 Electoral College votes needed to be elected president.

While Nielsen ratings likely have not been used as a voter intent signal in the past, MediaPost requested this data because in recent years, conventional political polling data has proven increasingly less reliable due to respondent biases and other factors.

And while politicos have been modeling other data -- especially so-called prediction markets data derived from virtual betting odds -- its unclear how those models will hold up in this year's election given how many unprecedented variables, and rapid-fire shifts in the news cycle there have been.

In fact, in conventional Nielsen parlance, the Republican ticket couldn't have asked for a better "lead-in" to generate interest in its event. But even though the assassination attempt occurred just two days before their convention started -- and even though it happened inside the state of Pennsylvania -- the Democratic convention outperformed the Republican's in that state by 8%.

In terms of conventional political polling, Ipsos' election tracking team began segmenting polling for the seven swing states months ago, and its most recent poll, conducted just after Harris and Tim Walz were nominated for the Democratic ticket, showed a small advantage for the Democrats in the swing states, albeit a less than statistically significant one (see below).

In terms of prediction markets data and models, the market data has also been trending in favor of Harris/Walz, and in an op-ed column in Thursday's New York Times, presidential predictions authority Allan Lichtman made his 2024 popular vote call for the Democratic ticket.