VoidSurf Profile picture
Oct 7, 2020 11 tweets 4 min read Read on X
We can't understand the present if we don't understand the past.

In this thread, you'll see what the past 120 years looked like for the USA. The source data is all available on the CDC website, though it was a real pain to aggregate. This chart shows monthly DPM since 1900: Image
These are the values I used for 2020. They were implied from weekly counts retrieved on Oct 3, so the vast majority of deaths should be registered for Jan-Aug. Future adjustments shouldn't impact the overall picture. Image
The red line you saw is called a LOESS regression. It's like a moving average, but more robust and tracks better with the center. It measures our changing expectation of mortality. Here we see only the positive differences (residuals) between the DPM and the red line: Image
This chart shows the residuals from the previous chart, but summed for each year from Aug till next July (12 months). Most bad months happen in the winter, so these values reflect how severe a season has been. Image
Here we see deaths/million for each Aug-Jul period. Roughly, this reflects the chance of the average American dying during the period. Note the alarming climb since the 2008 financial crisis - it's the only time there's been an upwards trend for at least 120 years. Image
These are the positive residuals from the previous chart after removing the overall trend. We can see by how much a period has exceeded the trend. The meaning is very similar to that of the chart in tweet 4. Image
Adjusting for population doesn't give a full picture. Age adjustment shows what each year would have looked like if it happened to a population with the same age profile. I use the population of 2000. We see a very different picture to that of the crude death rates. Image
Monthly figures aren't available by age group that far back, so I made an estimate by weighting by the proportion of all deaths in each month, per year. It's not perfect, but it should be close. Conditions have improved dramatically since 1900. Image
Finally, I'll sum the values from the previous chart for each Aug-Jul period so we can see what the known half of 2020 looks like relative to the past. Image
I hope this information gives you some perspective on what we're dealing with today. The conditions our ancestors dealt with daily were much harsher than even the worst of Covid-19.
We need to confront hardship with their strength and make sure we don't create horrible unintended consequences. I fear it may be too late for that, but it's not too late to change course and adapt to changing circumstances.

• • •

Missing some Tweet in this thread? You can try to force a refresh
 

Keep Current with VoidSurf

VoidSurf Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!

PDF

Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @VoidSurf1

Oct 25, 2020
We can't understand the present if we don't understand the past.

This thread will show you the history of death and disease outbreaks in Sweden. Source data was all retrieved from Statistics Sweden. Here are monthly DPM since 1851:
(repost - caption typo fixed + improvements) Image
These are the values I used for 2020. I've tried to be conservative - where there's a chance deaths might be understated, I've methodically adjusted them upwards. Also - I started this project over a month ago, so newer data is now available. These are still good though. Image
Winter has the most months with high mortality due to disease and harsh conditions. So, to make annual figures, I'll be summing from July to June whenever possible. This will give a good idea of what a season looked like. Image
Read 21 tweets

Did Thread Reader help you today?

Support us! We are indie developers!


This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Don't want to be a Premium member but still want to support us?

Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal

Or Donate anonymously using crypto!

Ethereum

0xfe58350B80634f60Fa6Dc149a72b4DFbc17D341E copy

Bitcoin

3ATGMxNzCUFzxpMCHL5sWSt4DVtS8UqXpi copy

Thank you for your support!

Follow Us!

:(