I spoke with Michael Starr at the NY Post today about my earlier item on the fact that there are now more TV sets in the US than there are people.
He asked me what I thought had caused that to happen. In the moment, I didn't have anything worked out nicely to support my guess, but as Robert knows I am often wrong, but never in doubt. I guessed it was the twin trends of declining household size (for a variety of reasons), combined with increasing television ownership, likely because the inflation adjusted cost of the average television set has gone down.
After I was off the call, I found the numbers and put together this chart. Lucky for me the trends looked like I hoped they would. As it turns out, we crossed the "more TVs than people" threshold in 2005.
Although I still don't have any information on the inflation adjusted cost of the mythical average television set, I'd love to hear from someone who does.
All the data is from the latest Nielsen Television Audience Report.
As this data is from Nielsen Media Research, I assume it’s from the metered households? If so, there are probably NOT more TV sets than people in the US because of the relative scarcity of sets in college dorms, retirement homes etc. that are not at all or not fully measured in the panel.
Really interesting question. Other trends to consider:
1. Easy credit fueling a boom in purchases of home electronics
2. Later-generation TVs lasting longer. For example, I have an analog set purchased in 1990 that still works perfectly.
Ray, I have to disagree with your points. College dorms are part of the Nielsen sample. Also, the Nielsen data excludes the literally millions of TVs that are in hotel rooms, bars/restaurants, workplaces, gyms, classrooms, etc.
In reality the gap between number of TVs and people in this country is probably far greater than what is captured by Nielsen.
These numbers aren’t *from* metered households but likely gathered before and then used to *create* the panel of metered households (i.e. Nielsen Families) to mirror the US TV population (as best they can).
Mikey, I know that dorms have been introduced into the sample, but I don’t think that they are at the representative level yet. I concede your point about public TV screens; I was thinking of personal-use sets.
Bill, I find it hard to believe (but don’t know) that Nielsen is using “number of TV sets” as a weighting variable for its panels, because it’s just too cumbersome, and if they don’t, it makes no sense for them to screen for it. So if the metered households are unweighted for number of sets, the sheer size of the combined panel should make it perfect for mining for exactly that sort of data.
(In my experience, it’s amazing how many people will let you put a meter in their living-room but balk at the installation of meters on secondary TVs, so it WOULD make a lot of sense to give greater weight to viewing on secondary sets, but can you imagine the added layer of complexity?)