Comment by elictronic
1 day ago
Demographic information is useful for medical, financial, educational, and so many other items.
The current admin doesn’t need it to discriminate, you can just access cameras and license plate readers and target easily that way.
The purpose is to scare people into misstating or obscuring data to reduce total house representation for an area. It’s to win votes, there are much better ways to do all these things than use this data, but effecting the vote with limited impact is a huge money savings.
The whole thing is over noise introduced to publicized documents to disguise the specifics of people in specific addresses, and malicious actors misrepresenting that as fraud.
For example census data was and probably is available on the block level but in order to avoid exposing the data of people living in these blocks that might be a few families, the publicized data aggregated and smoothed the data over blocks so you had cases where a block with a few single family residences reported over 100 people living in them. Obviously certain actors shouted voter fraud over the top of their lungs.
So now the law says no fudging of publicized data to preserver privacy, the government always had the actual accurate data.
The obvious solution to this problem is to just hide the sensitive data instead of fudging them.
In the above example the block would now report 10 people living there, but not their racial, religious, ethnic, or socioeconomic conditions.
> Demographic information is useful for medical, financial, educational, and so many other items.
What does this have to do with the census? A doctor would know the race of their patient without needing to deduce it statistically from their neighborhood.
Also, don't we not want financial institutions using demographic data decisions in making loans etc.?
How do you show that financial institutions aren't using demographic data unless you have it and run statistical analysis that shows they aren't? Banks aren't the government and can use methods that are illegal for the government to do in order to collect data. Let's say a bank manager is racist in giving out loans, how do you prove it unless you have the data to show it?
How do you know if they are even if you have statistical data? Race correlates with all too many things (e.g. education, income) that affect the default rate, so a statistical disparity is expected as a result of the things that correlate with race and affect defaults even if the loan officer isn't penalizing race at all.
Statistics are doubly useless in that context because a given loan officer might process something like 20 loans a year, which is too small a sample size for statistics to show anything with high confidence anyway.
The way people like that get caught is when they incriminate themselves. And the real way you solve that problem is by ensuring a competitive market for loans, so that the economic inefficiency of not giving loans to people who would pay them back actually negatively impacts the institutions that do it, and the borrowers can find numerous other institutions willing to do business with them, instead of the racist loan officer both being the borrower's only option and their bank not suffering consequences from it because the lack of competition allows them to stay in business by also overcharging the customers they accept.