This is where we get extra snazzy and graph our data array. Patterns get very obvious when you start drawing pretty pictures. Note to open source software users. The spreadsheets in Open Office are unusable for this sort of basic graphing, it will assume extra decimal places when it renders graphs or charts, producing completely wrong data in the graphs. It's not actually fixable either manually or problematically. It's just a crappy app. I recommend Gnumeric. It not only works, but is actually intuitive.
Let's start with AM radio. The general consensus is that it is in a slow decline.If you look below that's true. The AM band has lost 200 stations in about 20 years. That's only an average of about ten per year, less than one per month. I find this data unsurprising and therefore uninteresting.
Next is FM radio. FM began to match AM radio in ratings before 1980. My gut feeling is that FM is today where AM was 30 years ago. Growth is slowing, but it's faring better than AM by far. I'll compare the two in this chart.
You can see here that AM and FM bands actually cross in column four aka 1994. After that the growth of FM tapered, but continued. In other words it's rate of growth is diminishing but it is still growing. So while AM is closing stations, others are opening FM stations. Nothing surprising here either.
Above is a chart for the growth of LP FM. It too has no surprises. It began rolling out in 2004 after some fits and starts. It was ultimately rescued from the malice of the NAB by U.S. Senators John McCain [R], Maria Cantwell [D], and Patrick Leahy [D]. Thank you all. So what you see here clearly is that it didn't exist until 1994 and then rockets up to what will probably be a bumpy plateau. The truth is that they only began including that data in the summary reports that year so there should be a bit more of a ramp on the front end of this line. It's also important to point out that this was tempered in licensing by the Great Translator invation of 2003. More here. 13,000 translator applications were filed in March of 2003. This was an organized effort by large-scale religious broadcasters to block out RF real estate while it was still possible. For the most part they succeeded and many small markets got satellite-fed religious programming instead of local content.
The general public still listens to radio. The ratings book tell us that while tastes are diverse the general categories of popular radio stations are music, sports or news. A radio dial crammed with religious chatter is not what the public wants. There is definitely room for some religious programming on some stations in every market. But what graph indicates above is not appropriate. They are effectively spamming the radio dial, and as a result they will drive users away from the platform. In other words: this is part of the problem.