Demography obviously has the power to influence many aspects of culture. The question is, in a given case, is the purported demographic change true? If not, then we're done. A non-existent change cannot have caused other changes. Recently in a thread about the decline of musicals since at least the 1950s, Whiskey offered a demographic explanation:
It's not the cult of authenticity that killed musical innovation, it's demographics. Not enough White teens to both provide innovators, and the market for it.
From memory (I won't bother looking them all up), he's offered this for a lot of changes between the 1950s and '60s vs. today -- not as many white teenagers. Is this true?
Obviously there are fewer teenagers now than in the 1960s because of the Baby Boom, but his argument is still wrong. That is because the period when teenagers and early 20-somethings made up the largest chunk of the population was not the 1950s or the '60s at all -- it was around 1980. As someone who reads a bunch of this stuff for graduate school work, I stupidly think that it's part of anyone's background information who talks about demography. So let's look in just a bit of detail -- actually, you only need to see two graphs.
First, let's look at the fertility rate over time. Notice that the peak is in the late 1950s and is still within a few percent of the peak through 1960; only after 1960 does it tumble downward. By the time the cohort born in this peak period reaches high school or college, it will be the late '70s or early '80s.
Next, let's look at the age pyramid in 5-year intervals. It's animated without a pause feature, so you may have to sit through it a few times to catch it, but the year when the 15 to 19 and 20 to 24 year-old bars are at their widest is in 1980. This says that our prediction based just on fertility was right -- those born during the peak around 1960 did not massively die off, and so young people were greatest in proportion around 1980. (You can see these pyramids in any good book on demography in your library.)
If such a puny fraction of a smaller population size was sufficient to make musicals and musical innovation marketable through the 1950s, then surely a larger fraction of a larger total population would make them even more marketable in 1980. Except musicals had been dead for decades by that point, and there weren't a whole lot of fundamentally new musical styles. The mistake in the argument is locating when young people made up the greatest part of the population. For the decline in musicals, there must be some non-demographic explanation. 
This also shows that demography isn't as all-powerful as we think in a broader sense: the period that we associate with youth rebellion, Hear the Voice of a New Generation, etc., is roughly 1957 to 1969 or so. But if it were sheer size that mattered most, we would look to 1980 as the golden age of teenagers. Not that there isn't something there in the popular awareness -- punk, new wave, Fast Times at Ridgemont High, etc. -- but we certainly don't think of the new wave era as quintessentially youth-dominant as we do The Sixties.
And it only gets worse if we look at the makers of youth culture -- they're not 15 to 24, but usually in their mid-20s. And 25 year-olds were represented the strongest in 1985, decades after The Beatles released Help! If sheer size mattered, we would look at "college rock," which exploded in the mid-'80s and was created mostly by people born in the 1958 - 1964 cohort, as the pinnacle of youth-dominant music-making.
Instead, it is the culture of the cohort that went through an intense hysteria that has persisted for the longest, rather than the culture of the cohort that was biggest in size. Going through a massive social hysteria forces solidarity on socially desperate young people (which is to say, all of them), and more solidaristic people will hold onto their culture for longer. That's why you still see Baby Boomers who look like they've just gotten out of a Weathermen meeting, and whose speech makes it sound like they're organizing a student sit-in. The same is true for Gen X people who still look like they just left a womynist slam poetry reading in a Berkeley coffee shop. They've preserved these things more because they feel a stronger connection to their generation.
The silent generations, on the other hand, never had to go through a hysteria and so don't feel such a strong affinity for their age-mates. No one born in 1959 still wears clothes that they would have at Studio 54 or CBGB's, and their disco-punk-era mannerisms and slang words are gone too. I can assure everyone that my generation, when we're in our 50s, will not still be wearing those retarded carpenter-style jeans of the late '90s, or blasting blink182 or Britney Spears out our car windows.
But I'll wager that in their middle age years, the Millennials -- who will go through a hysteria within the next 5 to 10 years -- will complain that rock music all went downhill after My Chemical Romance and Fallout Boy, will still be clinging to the skinny jeans and ballet flats look, and will still use "fml!!!" in their Facebook status updates.
In sum, as far as cultural production and persistence goes, sheer size of the group matters less than the strength of its bonds. Cohorts with a strong sense of belonging to a Generation will produce more and preserve it for longer, considering it something sacred that you just don't throw away, no matter how smelly may have gotten. And it is a generalized hysteria that brings socially defenseless young people together in tighter-knit groups.
 In Arthur De Vany's excellent book Hollywood Economics, he shows that Hollywood consistently ignores the niche for G, PG, and even PG-13 movies, even though they have higher return-on-investment than R movies. Maybe Hollywood people prefer making edgy R movies over cheesy G movies -- or whatever -- but there is a huge void there, and musicals would surely fall under this untapped niche.