As reported here: Report: Some census takers who fudged data didn’t get fired | AP News
It seems that the Census workers may have fudged the data a smidgen to meet goals and gain bonuses…but there is another side of this story that I can tell because I was one of those Census workers.
We were working with names and addresses that were at least ten years old, and Those On High were determined to get as close to 100% compliance as was inhumanly possible. I was sent to apartment buildings that no longer existed, the same ones over and over and over again. I was repeatedly sent to houses that were torn down. I was repeatedly sent to streets that no longer existed. I was repeatedly sent to summer boathouses that were unoccupied because the season was over. I was even sent multiple times to a marina attached to a golf course and, although the manager of the golf course said in writing that no one resided there, I had to do a write up on every boat slip even if no boat was tied up there. We had to get a minimum number of positive hits per day but it soon became impossible to do so for obvious reasons. They even sent me to a church around the corner from where I live five times because they were under the impression that someone lived there and they were determined to get personal info from this (Holy) ghost.
As far back as I can remember (I’m 43) it seems like each Census has been touted as “the most accurate one ever.” The unspoken message behind that is “all those others have been inaccurate, but we’ll get it right this time.”
So having a certain degree of inaccuracy is, if nothing else, in line with history.
I think most of us who worked on the 2020 Census knew that there were a lot of inaccuracies. (There are bound to be some in every Census, but this one was worse.)
Partly due to Covid – we were coming out months later asking people about where they lived on April 1st – many places the residents weren’t there back then. And the Census address lists were pretty inaccurate (not unexpected, with ongoing construction over 10 years) – but they refused to correct them. I was sent repeatedly to a laundry room in the basement of an apartment building, where the Census insisted somebody lived. Other Census workers were sent there too. The poor guy living next door got hit up 6 times by Census workers (and he had only moved in during June). Despite entering this info, we were sent back again! Lots of similar cases like that. Sent to real-estate offices (not residences) and to empty, abandoned office spaces.
Also, the software on the smart phone devices was poorly designed. A couple of things:
- if you accidentally keyed something wrong, it was difficult to correct. Often you couldn’t go back easily. And very time consuming, which annoyed people who were waiting to get back to what they were doing.
- Design failed to include shortcuts to make data entry easier (and faster). For example, if you’re interviewing John Smithson, it’s pretty common that his wife Mary & kids Susie & Tommy all have the same last name. But there was no shortcut way to enter that – you had to repeatedly type in that same name 4 times.
But there was a political trend to this – the administration in charge at the time did NOT want an accurate Census – it would do political damage to them. Everybody knew that an accurate Census would show an increase in the population in cities & suburbs (which vote Democratic) and a big decrease in the rural areas (which vote Republican). So no real incentive to be accurate. Also, I think it likely that most of the incomplete data or inaccuracies were in the cities & suburbs. The Census leaders certainly seemed to concentrate more effort into the rural areas than the population figures would justify. Rather disappointing to see the Census have become a political exercise.
I was sent back to the middle of a highway twice to get info from the imaginary owners of a house that was there before the road was built.
Breaking news from 1791.
Sorry to interrupt your day.
Whoosh.
Perhaps-wouldn’t be the first time. Please explain?
It was just a joke, following Bootb’s post about the inaccuracy of previous censuses.
Got it. 
I was hired by the Census, and resigned quickly when I hit a roadblock in training, specifically a block of videos depicting scenarios I might encounter, and to be exact, the one where a man answers the door waving a baseball bat, and what do I do about it?
(Answer: Mark the property as “dangerous” and move on.)
I depends which way the inaccuracy went. If there was a city address that had been torn down since the previous census, but the current census keeps trying to find someone there, that would go against your premise. If you were trying to exaggerate the rural population, you’d be happy to accept that the building (and the people in it) were gone.
If all the extra effort described by the OP only took place in rural areas, then you’d be right.
“censi”?
At times it was totally censless.
if i remeber the 1890 one was so off in some places due to political machines like Tammany hall and others of its ilk they had to get a prototype computer to untangle the mess
@Czarcasm, sounds like your managers were constantly sending you after households that didn’t exist.
Conversely, were you ever aware of the Census Bureau, through bureaucratic incompetence, ignoring households that did exist?
Since we were given computer-generated lists sending us to specific addresses, I had no idea if other workers were hitting the houses that I wasn’t sent to.
My wife worked the Census and had a steady stream of tales of woe and frustration whenever she got home from work. Sometimes it was the crappy software, sometimes it was non existent addresses or abandoned houses, sometimes it was uncooperative or maybe just plain skeevy people answering the door.
One day she got assigned an apartment complex that was still under construction, so obviously nobody was living there. And as @Tim_T-Bonham.net says, there was no easy way to flag the entire block of addresses as unoccupied - she had to enter a note for each one individually. And she couldn’t even copy/paste for some reason, she had to type out the same note for each one.
Apparently the GPS/map app on the government furnished iPhones was crap as well, but she was not allowed to use the Google Maps on her personal phone when she was having difficulty finding an address.
If it was summertime, late in the census gathering, it must’ve been really frustrating. They were trying to clean up addresses that hadn’t replied to multiple mail/phone/flyer attempts. They were grasping at straws, trying to get any kind of information, and sending you back to places (houses, shacks, vacant lots) that still weren’t “resolved”.
That’s when I got hired. And I loved being a spy… (gotta run, tell you 'bout it later).
Not even in Latin.
No, the problem was that it took eight years to complete the count by hand. They therefore held a competition for quicker methods and Herman Hollerith won with his electromagnetic tabulating machine.
Yes, understandable. But after multiple enumerators flagged that address as unoccupied, they still kept sending people back again and again, when it should have been ‘resolved’.