search

ryankemper / writings

This repo will contain various writings (articles, free-form, etc) as I iterate on them. When stuff is ready it will be published somewhere "official" (medium, personal blog, etc - haven't decided specifics yet)
0 stars 0 forks source link

Suicide numbers out by 20x #9

Open 0atman opened 4 years ago

0atman commented 4 years ago

I think you've missed an important part of the suicide analysis:

Facts Stated In Your Article

As you've done excellent research, I will accept all figures you've presented as-is:

That's an increase of 7.11x over last year.

In 2017, one in 6891 people committed suicide.

My Analysis

If unemployed people have double the suicide rate of employed people, that means that they are overrepresented in suicide figures by the same. Therefore, 1% (double the 2019 unemployment rate) of the suicide deaths in 2019 were by unemployed people.

Your article says 50,000 deaths caused by COVID-related unemployment, I don't think that's right. I think the number is 2878.

I think your mistake was thinking that if unemployment doubles, the suicide rate doubles.

ryankemper commented 4 years ago

Thanks. To be clear the 50,000 was more a number I threw out to account for the other factors (isolation, culture of widespread fear, etc), but I think it would be much better to give the "hard" number you calculated and then have another sentence saying that that represents a lower bound and that we could very easily see far more.

Working on integrating this change.

On Thu, Apr 30, 2020 at 10:24 PM Tristram Oaten notifications@github.com wrote:

I think you've missed an important part of the suicide analysis: Facts Stated In Your Article

As you've done excellent research, I will accept all figures you've presented as-is:

  • w/e April 4, 2019: 1,757,864 unemployed claims (unemployment rate: 0.5%)
  • w/e April 4, 2020: 12,506,993 unemployed claims (unemployment rate: 3.8%)

That's an increase of 7.11x over last year.

  • an unemployed citizen is 50% more likely to commit suicide
  • 47,173 suicides in 2017
  • (population in 2017 was 325,084,758)

In 2017, one in 6891 people committed suicide. My Analysis

If unemployed people have double the suicide rate of employed people, that means that they are overrepresented in suicide figures by the same. Therefore, 1% (double the 2019 unemployment rate) of the suicide deaths in 2019 were by unemployed people.

  • In 2017: 471 of suicides were by unemployed people
  • 2020 has 7.11x more unemployment: Therefore 3349 of suicides will be by unemployed people
  • This is an extra 2878 people, if all the above holds.

Your article says 50,000 deaths caused by COVID-related unemployment, I don't think that's right. I think the number is 2878.

I think your mistake was thinking that if unemployment doubles, the suicide rate doubles.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/ryankemper/writings/issues/9, or unsubscribe https://github.com/notifications/unsubscribe-auth/AC7DOU4BE5J67DDSUVOMTVLRPJMHDANCNFSM4MW34LFQ .

ryankemper commented 4 years ago

BTW, I didn't account for the fact that https://www.dol.gov/ui/data.pdf, which I linked to, is regularly updated at that link. So there's now a new week of data to tie in.

0atman commented 4 years ago

Excellent, the bloggers I most respect somehow manage to keep their data up-to-date with many edits, to reflect an emerging situation.