Podcasts and Self-Selection: A hidden problem?

Credit:    Kaboompics

Credit: Kaboompics

As you may have noticed, I'm a big fan of podcasts. But this post is not another one of those fanboy posts telling you how amazing podcasts are. This is more of a random thought I had the other day about the nature of these wonderful pieces of audio. Hence, this is a "mini" blog post that is not necessarily connected to the regular material that I write, but that I still thought was worth sharing anyway.


Data Points

What got me started on this train of thought was when I started thinking about the characteristics of the people that listen to podcasts. I thought that since podcasts aren't so mainstream (yet!), the audience to listens to them must be the more tech-savvy and early-adopter type of crowd. So I did some digging and was lucky to find this recent study by Edison Research (which I believe I have referenced earlier) about podcast listeners. Although I was very surprised to see that the age of podcast listeners is surprisingly evenly-distributed, I was mainly interested in the socioeconomic data like income and education.

That said, here are the three data points that stood out to me from the study:

  1. A cool 35% of podcast listeners earn an annual household income that – wait for it – exceeds $75,000.

  2. Regarding education, 73% of listeners have done at least some college (including graduate and advanced degrees).

  3. Finally, 60% of listeners have either part-time or full-time employment.

These three points together lend a lot of credibility to the thought that got me started with this post in the first place: I believe that, for better or for worse, the audience of podcasts is very self-selecting.

Note: I focused on the data of the “U.S. Population 18+” although you could also choose to emphasize the data for the “Monthly Podcast Consumers 18+”.


Before I proceed, let me just say that when I mention the term "self-selection", I was inspired by the self-selection bias in statistical sampling. However, in this post I do not discuss it in a way that is exactly equivalent to the way it is used in statistics. Here I'm more interested in the people that (choose to) consume a good – podcasts – rather than the people that make up a sample for the purposes of a research study.

Anyway, it seems to me that because podcasts are not very mainstream yet and are a bit "trendier" than regular forms of media, the type of people that listen to them are, accordingly, early-adopters, more tech-savvy, more educated, and probably more in "in the loop" with current trends in the blogging and technology space. This is why I say that the audience "selects itself" based on the characteristics of the product.

Now you might think that this is logical, right? Many products in the world are made for certain types of people and thus only the people that are the right "fit" for those products buy them; running shoes are for sporty people that care about running, gaming consoles are for people that are interested in videogames, and Zumba classes, rumor has it, are for people that care about having a ridiculously toned backside.

However, I don't think that this situation is different for one basic reason. With over 250,000 podcasts out there, almost every topic under the sun is covered, meaning that people that like tech should be listening to tech podcasts; people that like psychology to psychology podcasts; people that like economics to economics podcasts, and so on.

What I'm trying to emphasize here is that the nature of the podcast 'product' itself is really quite generic and not so radically complicated: it's simply on-demand radio. Given this generic nature, I would have thought that it would easily become a mainstream product and thus capture the masses of the population very quickly. But this research suggests that this has not quite been the case as we generally see employed, well-educated, and more affluent people consume podcasts.

Note 2: Even though I talk about this study as if it speaks for ALL podcast listeners, I have to point out that it “only” accounted for 2,000 listeners (which may or may not be a representative sample). Also, the study used gathered data from telephone interviews; I’ve always had very strong doubts about telephone survey data for research purposes. That said, Edison Research is the only organization that I could find with truly comphrensive research on macro-level trends in the podcast industry.

Although this is a biased example that I'm probably conveniently using to illustrate my point (although, in my defense, it's very difficult to find data on this topic), I still think it's worth pointing out. The characteristics of listeners of my favorite podcast, The Art of Charm, seems to tell the same story that I highlighted above from the Edison Research study, namely that, on average, the listeners are employed, very affluent, and highly educated.

A hidden problem?

What's the problem then, you ask? The thing that worries me is that the people that truly need this high value, free information don't get/use it. Less affluent people, those who haven't been fully included in or able to benefit from the education system, those that have fallen behind and thus have most to gain from this; these are the people that need it most. And a quick look at some of the research shows that there is quite some promise regarding the educational value of podcasts.

Am I advocating for a sweeping governmental policy implementation that champions podcast adoption all over the nation? That's certainly too dramatic. I do however think that access to information is unquestionably empowering and, thus, I'm worried/disappointed that the people that need this information most are not getting it. Could it be because these people don't have access to the necessary technology (i.e. smartphones, primarily)? Who knows...

Perhaps podcasts won't radically transform people's lives, but I think it's a terrible waste that their benefits are not shared as widely as they should/could be.

See you, Space Cowboy.