Population Estimates - 2019 vs. 2020

[warrtalon]

Hello, @acharney2 , we discussed this previously, and you mentioned the population numbers were being updated.

Are you able to share the timeline for that? I'm asking again due to the urgency of this issue for how it impacts POC, children, and elderly.

When using 2019 Intercensal, the results are non-sensical and make it impossible to target policies to the proper demos:

• 18-24: 101.85%
• 35-44: 100.04%
• Asian/NHPI: 102.38%
• NA/AN: 119.74%
• Asian Adults: 108.7%
• NA/AN Adults: 133.43%

I know there is a disclaimer saying small groups may go over 100%, but that would only apply to NA/AN. The Asian/NHPI, 18-24, and 35-44 groups are huge. This indicates a major denominator issue.

The 75+ groups also indicator a major problem, but we can't know if it's a data problem or a real problem with that demographic. This is how extreme the numbers are:

• More 75+ Unvax than all of 18-24 + 35-74 combined
• Equal 75+ Unvax to all of 13-24 + 35-74 combined
• Both 75-84 & 85+ have more Unvax than other age groups besides 25-34 & 5-12 despite having lowest pops

When you combine this with the published Death rates per age group (Vax status not provided) where 75+ death rate is as much as 250X higher than 35-44 (infinite for younger), it leads to a misunderstanding of what drives the Vax vs. Unvax COVID risk ratios:

When applying 2020 census to the Population Vax%, the overall trend looks like this through week of 1/1:

This 2nd version goes through 1/29, but Vaccine Record Matching is not complete after 1/1:

I know this is a lot, but these details are crucial for Commissioner Chokshi and Mayor Adams to make well-informed decisions regarding public health policy.

Thank you! -Clayton

[acharney2]

See here for the Census schedule for releasing population data. We won't be able to update our estimates until after the Census releases data at the level needed: https://www.census.gov/programs-surveys/popest/about/schedule.html

[warrtalon]

See here for the Census schedule for releasing population data. We won't be able to update our estimates until after the Census releases data at the level needed: https://www.census.gov/programs-surveys/popest/about/schedule.html

Ahhhh, I see. That's helpful.

To be sure I understand:

• There is some newer data available as of 7/1/21, but it's only at the National/State level
• The Metropolitan area population and demographic components are expected March 2022
• But there are other details like breakdown by age/sex/race that come even later in the year

Due to this, it means:

• NYC Health only has 2019 Intercensal available at the demographic level. This uses Vintage 2020 Population Estimates, which do not incorporate nor consider the 2020 Census results (similar to what you told me already)

But since we know the 2019 Intercensal is way lower (~500k) than the 2020 Census, can you not at least:

• Use the 2020 Census for the overall Total Population Vaccinated % calculation?
• Use the 2020 Census to determine the Unvaccinated population denominator?

Thanks again!

[acharney2]

Thanks for your input and patience. Our protocol is to use consistent population estimates across all calculations for ease of comparison. We plan to transition to the 2020 Census once the data are ready.

[warrtalon]

Thanks for your input. Our protocol is to use consistent population estimates across all calculations for ease of comparison

But this then knowingly overstates the Vaccinated Population % while knowingly inflating the Unvaccinated COVID rates. Right?

This seems critically important for the public and political leaders to know the true numbers. Skewing all in one direction is detrimental for everyone.