Not a historian but I have a degree in Library and Information Science. Microfilm is just the middle of a continuum of micro/digital transfer that has been doing on for years. In one library I worked in we used to have "Microcards" which are from the 1940s and 50s. We have one machine left that could read them, but they are essentially just a piece of cardboard with the documents (in this case science articles) printed onto them in a very small but magnifiable format. After microcards came Microfishe. The problem with the cards was that even today it is very difficult to make a copy from them. With fische you could throw it onto the machine, locate the pages you wanted, and print them page by page. Problem is fische is kinda flimsy and can get damage/ lost easily. Fische is essentially just small pages of plastic, similar to the cards, with the benefit being it can be easily copied. However, if you loose a sheet its gone, and can be rather pricy to order a new copy.

After fische came microfilm. This was really popular around the 1970s. The benefit of microfilming was the it was study, copyable and compact. One of my friends who works in an archive still insists on microfilming, because of its beneficial storage properties. Microfilm made in the 1970s is still accessible today. You can take a microfilm from 1970, put it into the machine and its still readable.

Prior to microfilm documents would have been preserved on cards or fische, but after microfilm came out these two methods become obsolete. The reality is, in most research libraries that existed from 1940s onward you can probably still find large collections of fische (less with cards, although I knew a scientist who was reduced to borrowing a microcard and reading it with a microscope and copying the information by hand).

You might wonder why not digitize everything...well it comes down to a couple of reasons

1. Labour costs - its labour intensive to copy documents
2. Subscription costs - Many of the online stuff you use is not owned outright by the library, they pay a subscription fee for access to it. This is a whole other discussion, but why pay repeatedly for something you already have on the shelf
3. Data migration - think of how many times you have upgraded your computer system. Think how often libraries do. Think how often the software you use changes or the formats change. When I learned about this in Library School the rule of thumb was that you had to migrate data every five years.....I think its more like 2-3 now. I remember one librarian I knew complaining that the library was finally converting to Windows Visa from XP...about a month before Windows 7 came out....this was to avoid data migration problems moving data from XP directly to 7.

I can't say for sure why microfilming took off when it did, but I suspect a lot of it has to do with cost savings. In Canada a lot of the original documents were microfilmed and then the originals were discarded, as the information in them now existed in a smaller, accessible and more easily copyable format. The danger in this is that if anything went wrong in the microfilming process (missed a page, it turned out fuzzy) you can't go back and redo it. But in the 1970s this must have seemed like the best of both worlds, you could keep the information, make it easily accessible, and stop having to pay for storage of all these musty old books and documents.

When digitization came along in some cases all they could do was digitize the microfilm. Canadian genealogists run into this all the time and the census, land registration and ship manifest lists exist only in this format. On the other hand, at least there is SOMETHING left to access.

If I am remembering correctly it is estimated that if you store microfilm properly it should be readable in 200 years. For a librarian you can put it on a shelf and forget about it. Compare this to digitized material which need to constantly be migrated. Its easy to see why some of us still see microfilm as a very useful resource.

Sorry this was so long and rambling.

Quite a few reasons:

1. Half life of polyester microfilm is rated anywhere from “a few hundred years” buy conservative people to 1000 years by some people. That’s WAY better than paper or bitstreams. It does however get worn out by being used (scratched up) so there is usually an access copy and a preservation copy.
2. Tolerates both fluctuations in heat and humidity with grace. This means if you can’t afford even basic climate control in your archives stacks (yes, this is still a harsh reality for a lot of archivists in America) the microform can take the pressure.
3. Takes basically no money to keep it. Digital archives take lots of money in servers and electricity to keep. If you’re a cash strapped archives microform is your cheapest backup option.
4. Is very small. Being an archivist sometimes feels like a constant battle against the walls. Unlike say a non-collecting library (which is almost all libraries) we never get rid of anything. We are always growing.
5. Can be mailed very cheaply (via Library Mail) for Interlibrary Loan purposes. For one of our most important collections in the archives we have a full microfilm backup that circulates.
6. Very low risk of media obsolescence. You might have a few 3.5” floppies around your house, but odds are very low that you have a reader for them. Not only is microform reader technology still being produced, it’s actually getting better. AND, if the apocalypse comes tomorrow, a person of moderate to low intelligence can figure out how to access microform technology pretty quickly. (Flashlight + white wall in a pinch.)