It was a lot of people, in other words. And CNN's Brian Stelter reports that the number didn't drop off significantly over the course of the 90 minutes — meaning that nearly everyone who saw Donald Trump's comparatively effective first 20 minutes also saw the last hour, too.
But that "most-watched" figure is a bit misleading for two reasons.
First, the population has grown over the course of the past 40 years. There were 100 million fewer Americans in 1976, so the 69.7 million who tuned in to see Gerald Ford and Jimmy Carter in their first debate was about a third of everyone in the country. 84 million Americans in 2016 is only about a quarter of the total population. As a function of total population, using that figure, the debate was the 12th most watched.
Nielsen also estimates the number of households with televisions each year and the percentage of those households that tunes in to a program. As the population has increased, so has the number of televisions, generally speaking. Just shy of half of American households watched the debate, far fewer than the 60 percent of homes that tuned in regularly in 1960.
Which brings us to the second point. I said that the number of televisions has generally increased, but, of course, the ecosystem is changing. In 1960, the debates aired on the three major networks. On Monday, it aired on ABC, CBS, Fox, Univision, NBC, Telemundo, Azteca, PBS, CNBC, CNN, Fox Business, Fox News and MSNBC.
It also streamed online, including at Twitter. That streaming traffic is not included in the 84 million figure reported by Nielsen, meaning that some huge chunk of viewership isn't captured in that number. Nielsen reports that 2.7 million Americans "interacted" with the debate on social media, which doesn't really tell us a whole lot more.
It's a reminder, though, that the 84 million figure may be a smaller percentage of the population than for previous debates, but that it's also probably lower than the actual total. The number of people who watched the debate is somewhere north of that — probably not 10 or 20 million north of it, but almost certainly some seven-digit number.
The reason we track this viewership is in part to get a sense of what the repercussions will be. A heavily watched debate that featured a lopsided performance would be expected to move the polls to some extent, as the first debate in 2012 did.
Regardless of the precise total number of viewers, one thing can be said with confidence about the ratings: It's probably one time that Trump might wish fewer people had tuned in.