2. So how do you know that back in 1977 the Social Security Administration did not have some of those scanners, because, as I see it, if any federal agency could have used them, it would be the Social Security Administration, an agency that handles thousands of pieces of paper every day.
Okay, so I actually did a little research, and I can say with certainty that the Social Security Administration did not use scanners to process applications 1977.
The first optical character recognition scanner, that is, the first scanner capable of inputing written characters on written paper into a computer database, was the Kurzweil Reading Machine. Its initial purpose wasn't scanning data from documents into databases, but helping the blind read, though it could do the former too. The first prototype was built in 1975, and the first public demonstration of the machine was 1976. As of 1977, no government agency was using it only any large scale. Some government agencies were testing it, but to help the blind read, not process paperwork. You can read more about it here:
http://www.rehab.research.va.gov/jour/77/14/1/kleiner.pdf
The machine wasn't available commercially until 1978.
So it's pretty certain that in 1977, a clerk at the SSA had to type in application data by hand.
3. I'm trying to imagine the processing steps of a Social Security application sent from Hawaii back in 1977 after it entered the Baltimore office.
Your image is probably acurate. The application comes into the Baltimore office mail room. Most likely the mailroom clerk opens the applications and puts them on a pile for processing. Another clerk distributes them to the data entry clerks. The data entry clerks, likely sitting in front of a primitive computer screen, type the data into the a computer database. Once the data are entered, another clerk probably files them away for a time somwhere, and they are eventually shredded.
Why are you so intersted in all these details about clerical work at the SSA?
I'm not buying the transposed numeral idea due to the fact that checking for mistakes is second nature to anyone that used OCR on a daily basis, and, unless a complete noob, would most likely know which state which zip code belonged in.
Of course, if they didn't use OCR at SSA, then the above is moot.