(
From the diaries. Interesting stuff -- kos)
I'm sure everyone is wondering if AirAmerica is succeeding or failing. Well, I've done the research and I can tell you the answer. Numbers and graphs inside.
From rightwing sources you hear reports of stations that are doing badly. WLIB is a favorite example, because it is in New York, a "liberal stronghold". From leftwing sources you hear reports of stations that are doing very well, such as KPOJ in Portland, OR. But a reality-based community has to get past cherrypicked data and find out what the overall picture looks like.
(Disclaimer: I am not a statistician. That said, I think my methodology is fairly obvious and sound.)
My original plan was to take AAR's list of stations and Google to find the ratings data, hoping to dig up a few of them someplace. As it turns out, you can get a lot of ratings data, including historical data for the past year, here. This site is the source of all my data.
I created a spreadsheet that listed the stations and saved the ratings information from the above linked page. Unfortunately AAR doesn't note when a station joined them and I only started this project a few months ago, so my spreadsheet in some cases claims listeners in past quarters that weren't really listeners then. Since that overestimates previous quarters, though, that only serves to underestimate the growth curve, so it's a fairly conservative measure. Also, not all data is out for the Spring 2005 quarter, but enough is out that I can project pretty much what the quarter will look like overall.
The first question I wanted to answer was if AAR was growing or not, so I divided every quarter's totals by the Winter 04 quarter (this is called "normalizing"). I found that there was a bump around election time and then a sag afterwards. That makes sense, but I wanted to know if the conservative radio market had seen the same thing to the same degree.
It was then that I realized I needed a baseline to compare AAR's growth to. I chose Rush Limbaugh. I chose him not because he's the most conservative or fastest growing or hottest property. I chose him because he's the most widely listened to (AFAIK) and would thus provide a good, stable baseline to compare against. So I found a list of stations Limbaugh is on and chose stations that were in the same markets as the ones I had AAR data for. I created a clone spreadsheet with Limbaugh/baseline data that works just like the AAR spreadsheet.
It is now possible for me to show how well AAR is doing relative to itself and relative to the rest of the talk radio market (at least in those cities).
First, the bad news. Mainstream (which is largely conservative) talk radio has many, many, many more listeners than AAR. For the selected cities, the overall listenership of Limbaugh/baseline is about 3.8 million this quarter as opposed to about 1 million for AAR. And don't think to yourself that AAR has 1/4 the listenership of Limbaugh, because Limbaugh is on in many more cities where I didn't even collect data because AAR has no presence there. That said, Limbaugh's 3.8 million is down from 4.4 million at the Fall peak whereas AAR's 1 million is up from 800,000 in the Fall. Here's a graph of absolute numbers:
(Y-axis labeled on the right side)
All that vertical whitespace is growth that AAR needs just to catch up to Limbaugh in the markets where AAR even exists. You can see there's a lot of work to be done. However, you can also see two other things.
- AAR is gaining.
- The Limbaugh baseline is falling.
This growth/decline can be seen more easily using the normalized numbers. Here each quarter for each dataset is divided by the first quarter for that dataset. The resulting number is basically a growth multiplier for that dataset since Winter 04 (just before AAR was launched). For instance, a Y value of 1 means that at that date the listenership was exactly equal to Winter 04 while a Y value of 2 would mean it was double.
(Y-axis labeled on the left side)
As expected, the graph slopes upwards to a peak during Fall 04, when the election was held. Then a decline as the excitement fades, though not quite so much for AAR. Then in Spring 05, while Limbaugh/baseline continues a downward slide, AAR takes off like a rocket.
Some may say that it is misleading to choose to graph the growth instead of the absolute numbers. However, I think it's a fair method in thise case. It is already known that AAR has many fewer listeners than Limbaugh--he's been around for 15+ years while AAR started fewer than 18 months ago. The question is, who is growing. These graphs show the answer is that liberal radio is the one that's growing.
(If you want something you can quote, here's some percentages: The mainstream talk radio market, represented by Limbaugh, is down 1% since the Winter 04 ratings while AirAmerica Radio is up 32%.)