I was doing research a few years back to finish out my bachelors, and I had the option of using microfilms and of using database searches from newspapers-for a fee. Now, my research was from something from the early 60s, so I could dominantly use microfilms, and had hours of fun reading the collateral stories, as well as get a few great points for my paper, just by reading the headlines of the newspaper, while my story was on page 10. If I were to have used the database method exclusively, I would have been denied the benefits listed above. I do know that some libraries dispose of some old microfilm files, and I worry that this will extend to newspaper files, in favor of databases.
So, question being: Are newspapers (and periodicals) still, routinely, microfilmed and stored as in days of yore?
Current newspapers are routinely microfilmed and the films then made available through city libraries here in New Zealand. If we’re doing it still (and not only holding onto the old ones, but making copies), then I can’t see why it would be very different elsewhere.
The newspaper for which I write has its entire content microfilmed back to 1859 and continues to do so.
Yep, newspapers are still doing it. There’s a lot of value in being able to see the page, not just read the text, and I imagine that digitizing whole pages is still too expensive and clunky compared to the relatively simple microfilming process.
Doesn’t microfilm degrade badly over time? I thought there was a huge crisis because of decaying newsprint and worse microfilm.
Of course they still microfilm. Digitization is expensive and proper indexing requires actual people, so you’re not going to see a lot of it for the Middle of Nowhere Times, Messenger, Banner, and Press. As for durability, microfilm silver halide masters are pretty good when stored properly (certainly a whole lot better than newsprint) and they don’t require any software, hardware, or anything else to look at. You could discover them in a post-apocalyptic storage room and figure out what they were and how to read them if you have the sun and a magnifying glass, which is IMHO a big plus.
We have these new fancy readers that scan into image or PDF files, which is pretty cool.
Zsofia addresses the issue well, but I wanted to add something to perhaps clear up the point. The masters can degrade over time, as all film products can, but proper storage can provide a lifetime of many decades, maybe even centuries. And of course, after 80 years you can produce a new copy of the originals and start the clock over again.
The prints taken from these masters suffer over time as they are handled, especially those open to the public who may not be thinking of preservation. These working prints can get scratched or torn, but these are only copies. If my library’s reel of the NY Times from 1903 gets ruined, they just order another one.
The huge advantage of digitizing newspaper collections is that they can be searched. Back in the 1980s when I was doing research for the baseball encyclopedia, it wasn’t unusual for me to spend a day in the library going through a years worth of a paper day-by-day, looking for some mention of a particular player… to explain why he was injured, or to find his exact date of death, or some other tidbit of info. I’d sometimes spend days getting bleary eyed and coming up empty, wondering if maybe I skipped a single page from April 11th that might have had the nugget I was looking for. Now, I can type a player’s name into a search box and get results – every time his name ever appeared in a wide range of papers. I can do in ten minutes what used to take me a month.
The downside of digitization (or even microfilm) is that you often end up with errors in the process, pages that are folded over, torn, or askew when scanned. And as Nicholson Baker wrote in his book “Double Fold,” there is an unfortunate trend to destroy the originals after the scanning is done.
And to get back to the OP’s question, yes, they still microfilm newspapers. The major producer is called University Microfilm.
We’re considering getting this new fancy reader that doesn’t “touch” the film - in other words, there aren’t those two glass plates for the film to go through. That would increase the life of our film. The funny thing is that the more access you provide, the more people use the film and the more beat up it gets and etc., etc., etc - we’re doing this massive obituary indexing project, and have compiled front page death list notebooks so you can find your dead people if you know roughly when they died (from SSDI or the South Carolina death indexes) and our film is jacked the hell up. We replaced most of it a year ago and it’s already showing wear. That’s what you get for giving the people what they want. 
One thing to remember is databases may not pull correctly. I used the Chicago Tribune index and I found that if I tried to pull information, a lot and I mean A LOT of times, it just doesn’t pull up. But if I go to the actual paper and read it online, I go through page by page, I find the information.
I guess the database isn’t set up to pull every word in the paper. I would find it interesting, that I could not find even major crimes by pulling a search and but by saying, “OK the crime occured on this day” and just start reading the paper, I found many pages relating to that crime.
That’s the thing - without proper indexing (and I don’t just mean keyword searching) databases can be extremely unhelpful. A very common problem with newspapers that I deal with every day is that, obviously, the article in the paper may not have the name of a victim, because it hasn’t been released yet, or the name of the perpetrator, because they haven’t caught him yet. Granted, not many newspaper indexes (even the hand-done ones) are good enough to do that, but it’s the ideal.