More on the Mystery Death Spike of CT and NC (that is totally bogus)

I will keep poking the CDC until they fix it

As a followup to last week’s post, I figure I may as well make a video while the CDC’s error persists:

I plan on continuing to poke the CDC until they fix their error.

I would prefer they just stop modeling unreported deaths in the first place. But if they’re going to model, they should be noting when they’re getting results that don’t make much sense. Otherwise, why are they modeling in the first place?

It seems like a waste of resources.

Example of reporting lag change effect

Here is a screenshot of a simple example of how one estimates unreported deaths:

Now, that’s an example where deaths used to be reported more slowly and now are reported more rapidly.

In that case, current estimates would overestimate how many deaths there are — that’s what is happening with Connecticut and North Carolina.

Old complaints of Connecticut and North Carolina

As mentioned in the video, I had complained on the blog over the dilatory nature of North Carolina and Connecticut’s death statistics reporting during the pandemic.

I found one key blog post from February 2021: Geeking Out: GDP By Country and Recent COVID Mortality Data

I did finally take a little time to find out why both North Carolina and Connecticut suck so bad in getting their death totals to the CDC (and I guess that’s why it takes a year to get finalized death stats, too. Dammit, CT and NC.) At least I know NC is working to get on the electronic reporting system that EVEN SOUTH CAROLINA IS USING, DAMMIT NORTH CAROLINA.

I still have no idea re: Connecticut. Connecticut should be ashamed of so many things.

I don’t think I ever found out why Connecticut was so awful. I think Connecticut falls down on many categories of government.

I found this paper from January 2021, which confirmed that the worst states for reporting death statistics were Connecticut and North Carolina.

Results

On average, we found that all-cause mortality counts take 5.6 weeks to become accurate with less than 1% increases subsequently. We display the average number of weeks of delay for all 52 jurisdictions in Figure 1. In this figure, the slowest states were North Carolina, Connecticut, and Alaska, which are respectively 5.4, 4.8, and 4.7 weeks slower than average, and the fastest states are Maine, New Hampshire, and Vermont, which are 3.4, 3.1, and 3.1 weeks faster than average,
a gap of almost 9 weeks between the slowest and fastest states.

It can’t be speed or money that was the excuse.

In any case, both North Carolina and Connecticut have finally caught up in speed, from my own unscientific eyeballing of the datasets. Alaska is still somewhat slow from what I can tell, but it truly has some challenges that would be difficult to overcome given its geographic spread.

The CDC is the one that hasn’t caught up with reality.

Related post:

A Tale of Know Your Data: The Mystery of the Excess Connecticut Summer Deaths