Over the decades, there have been many occasions when the data published by the Census Bureau is either likely or absolutely in error. Typically, we rely upon national and state level population estimates to use as control totals. This always presents a dilemma for a company like AGS, as our primary goal is to create the best small area demographic estimates possible. On the other hand, divergence from the estimates of the Census Bureau will be met with incredulity by users. After all, what small private company would claim to have better numbers than the government of the United States?
In recent years, we have seen three massive leaps in the estimated population of the United States. The earliest 2020 Census tallies found nearly 2 million people more than the Census Bureau had estimated even as late as the summer of 2020, well after the April 1 date of the Census. Late in 2022, the Census Bureau added about 4 million people to its estimates. In December of 2024, nearly 3 million were added to the national population estimate, relatively quietly we might add.
National population estimates are relatively simple as there are only four components – births, deaths, immigration, and emigration. Assuming you have a reasonably accurate starting point, such as the 2020 Census enumeration, you can estimate the ‘natural increase’ of the population very accurately. Such estimates can be adjusted to account for anomalies in births and deaths from one year to the next. The main source of uncertainty is counting who is leaving or entering the country.
Over the past few years, news reports have indicated that more than two million people have crossed the southern border annually. The actual numbers may be higher, as estimates from the Department of Homeland Security account only for those detained and released at the border, as well as recorded instances of individuals evading capture (the ‘known got aways’).
The Census Bureau has the capacity to incorporate these figures into its national and state-level estimates. However, major revisions have been made only three times in recent years—each occurring in the month following a national election.
The latest revisions were reported, albeit without fanfare, recently on the Census Bureau site — https://www.census.gov/newsroom/blogs/random-samplings/2024/12/international-migration-population-estimates.html. These numbers were immediately reflected in employment and labor force estimates, and not without drawing commentary, such as Huge Upward Adjustment to Employment & Labor Force as Wave of Immigrants is Finally Included, Unemployment Drops, Wage Growth Accelerates: The Annual Revisions Are Here | Wolf Street.
To AGS, the primary question in this case is: in what census blocks are those five million people? We could add 1 person to each of the nearly 5 million populated census blocks, but this is clearly not very accurate. Many have no doubt been absorbed into the households of friends and relatives, but we also know that large numbers have been quietly relocated by the federal government into many small and medium-sized cities nationwide. What is known for sure is that the national housing stock has not expanded sufficiently to accommodate these immigrants – and it is the change in housing stock at the local level which is a key factor in knowing where demographic change is having the most impact.
Did AGS know that our national population totals were on the low side? Of course we did, since we pay attention to the news. But until these were acknowledged by the federal statistical agencies, it was ill advised to diverge that significantly from the national population estimates.
Our 2025A release in May will by necessity allocate several million additional people into the local estimates, using information gleaned not just from the federal agencies, but from careful tracking of change at the parcel/building level, along with additional external private delivery counts and consumer lists. But unlike the federal government, we won’t disappear our current release into the memory hole by hiding it nor will we replace published data with re-estimated data using the latest totals. After all, our goal is simple: to create the most accurate current estimates possible.
Recent Comments