Thursday, November 12, 2020

Lawler: Recent Slowdown in US Population Growth Accelerated in 2020

by Calculated Risk on 11/12/2020 03:52:00 PM

From Tom Lawler: Recent Slowdown in US Population Growth Accelerated in 2020

US population growth, which slowed significantly over the past few years, almost certainly grew at an even slower pace in 2020. All three key components of growth – births, death, and net international migration – should contribute to the slower population growth. Before discussing each, here are the latest estimates (“Vintage 2019) of the US resident population, as well as the estimated components of change, from July 1, 2015 to July 1, 2019.

 US Resident PopulationYearly Change% ChangeBirthsDeathsNet International Migration
7/1//2015320,635,1632,334,1550.73%3,992,3762,699,8261,041,605
7/1/2016322,941,3112,306,1480.72%3,962,6542,703,2151,046,709
7/1/2017324,985,5392,044,2280.63%3,901,9822,788,163930,409
7/1/2018326,687,5011,701,9620.52%3,824,5212,824,382701,823
7/1/2019328,239,5231,552,0220.48%3,791,7122,835,038595,348

So what do we know about each component of change since July 1, 2019?

1. Births: The CDC estimated that the number of US births during the calendar year 2019 was 3,745,540, down from 3.791,712. 2019 was the fifth consecutive year that the number of US births fell. While I have not seen any data on births so far in 2020, it is probably reasonable to assume that births over the 12 month period ending 6/30/2020 was about the same as in calendar year 2019.

2. Deaths: There is little doubt that deaths in 2020 will be substantially higher than in 2019, mainly by solely because of deaths directly attributable to Covid-19. While the CDC has not yet released the official “US Mortality” report for 2019, it has been releasing provisional weekly estimates of deaths for 2020, in part to estimate the number of “excess deaths,” or deaths higher than one would have projected based on trends, for this year. While the data are provisional and are not complete because of reporting lags, below is a tabulation of provisional death estimates by broad age groups that corresponds closest to the same calendar period as that used in the official Census population estimates.

CDC Provisional Estimates of US Deaths by Selected Age Groups
52 Weeks Ending:Under 2525-4445-6465-7475-8485+Total
7/2/201663,970129,045533,133499,219626,580836,0842,688,031
7/1/201765,147139,131541,372522,102647,667870,7252,786,144
6/30/201862,488140,377543,927540,434667,830883,4572,838,513
6/29/201960,072139,837536,890547,891678,081863,8632,826,634
6/27/202061,293156,400566,785593,714729,344919,6033,027,139

Note that the increase in deaths in 2020 is not just limited to the elderly age groups. In fact, in the largest % increase in deaths was in the 25-44 year old category.

While these data are by no means final, and the overall total for 2020 will probably be a bit higher in 2020, a conservative estimates for US deaths from 7/1/2019 to 6/30/2020 would be about 3.03 million, up by over 203 thousand from the previous year.

3. Net International Migration: Net International Migration (NIM) was estimated to have declined significantly over the past two years, at least in part reflecting Administration actions. The pandemic, combined with additional Administration actions, almost certainly resulted in substantial declines in NIM from February to June, but I am unaware of any data that would enable one to e/1/2stimate the drop in NIM. It is probably not unreasonably to assume that net international migration from 7/1/2019 to 6/30/2020 was no more than about 300 thousand. (Note that NIM estimates are subject to very significant measurement errors, for a host of reasons.)

If one assumes, births over the 12 month period ending June 2020 totaled 3,745,450, deaths over that period totaled 3,030,000, and net international migration over than period totaled 300,000, then the US resident population estimate for 7/1/2020 would be as follows:

US Resident
Population
Yearly
Change
%
Change
BirthsDeathsNet Intl
Migrate
7/2015320,635,1632,334,1550.73%3,992,3762,699,8261,041,605
7/2016322,941,3112,306,1480.72%3,962,6542,703,2151,046,709
7/2017324,985,5392,044,2280.63%3,901,9822,788,163930,409
7/2018326,687,5011,701,9620.52%3,824,5212,824,382701,823
7/2019328,239,5231,552,0220.48%3,791,7122,835,038595,348
7/2020329,255,0631,015,5400.31%3,745,5403,030,000300,000

For those who like to use “official Census population projections either to describe likely future demographic changes of the forecast other variables dependent on population total below is a comparison of the Census 2017 Population Projections (the last official long-term population projection) with the Vintage 2019 Population Estimates and my estimate for 2020.

 Census 2017
Projection
Vintage 2019
Estimate/2020
LEHC Estimate
Difference
7/2017325,511,184324,985,539525,645
7/2018327,891,931326,687,5011,204,430
7/2019330,268,840328,239,5232,029,317
7/2020332,639,102329,255,0633,384,039

As the table above shows, the Census 2017 population projections even before 2020 were way above the latest estimates, and that gap probably surged in 2020 to almost 3.4 million. This highlight why competent analysts do not use the Census 2017 population projections either to describe likely future demographic trends or as inputs into forecasts of other variables dependent upon population projections.