Based on the below links and estimates I will resolve no unless someone else brings me other evidence. Ideally dated to before the tweet was made.

@nic_kup to clarify: I will resolve no at the end of the 100 days if no other evidence comes in and I don't randomly stumble across more.

@nic_kup why dated from before the tweet?

@AlQuinn It isn't strictly necessary, but it is very easy for sources to cyclical references as people updates their articles to reference new articles. Makes it hard to know where the numbers came from.

According to an audit by the secretary of state in 2020 approximately 23,000 Medicaid records lacked SSNs out of their total dataset. But:

1. After accounting for valid exemptions (like newborns and foster children who are not required to have SSNs), only 1,980 individuals without SSNs remained questionable.

2. From a random sample of 66 of these cases:

• About 40% were appropriate (involved foster children)

• Only 12 individuals were found to be potentially associated with improper payments

• The report estimates between 162 and 558 total cases could be associated with improper payments

This makes me update downwards to maybe like 10%. In particular because the Do Not Pay agency seems reasonably sophisticated and willing to solve problems in a practical way.

Link: https://sos.oregon.gov/audits/Documents/2020-05.pdf

PS: Elon should expand the Do Not Pay Program which saves the government ~$286 for every $1 spent.

@nic_kup In Oregon (where there was an audit in 2020)  there were around 23,000 Medicaid records without SSN. But, 18,000 of them were individuals under 1 year old. After accounting for other valid exceptions there were 1,980 individuals remaining. Let us assume all of them are fraud (although report estimates only 20% after investigations)

Extrapolating to all of US (i.e. 80x) gives ~160,000 such cases. Average Medicaid spending is around $3,000 per person, but lets say all fraudsters extract around $30,000. That puts the national estimate at ~$5 billion.

It looks like in 2013(!) the "Do Not Pay" Program updated saying that there was $108 billion in "improper payments", although part of it were underpayments by the government. And it states that of that $64 billion were "actual loss to the government".

Note that it doesn't mention anything about individuals without SSN or TIDN. I will still update towards the statement being true, idk maybe 33% likelihood of being true.

Link: https://www.ignet.gov/sites/default/files/files/dnppresentation.pptx

PS: happy for someone to point me to newer numbers or correct me on interpretation.

What happens if you believe one statement but not the other? The first statement seems much more credible than the second.

@Kire_ The second statement is a lot more soft, i.e. someone credibly believed that x amount was fraud. I definitely put more weight on the first statement. I will also try and update as I come across evidence / reply to all comments with what I think.

Note that I asked o3-mini to find me any report giving a credible number before 2024-12-31 and it couldn't find anything:
https://chatgpt.com/share/67a8caa6-1914-8001-9dea-7bffa327bcec

You have an account on Twitter, that's all the bias I need to know

You have an account on Bsky, that's all the bias I need to know

@skibidist I paid M1000 that I didn't need to in order to find the truth. I am not betting on this market to stay neutral. Show me the evidence and I will change my mind.