Emails, Lloyd Potter, Texas state demographer, Sept. 28, 2015
From: Selby, Gardner (CMG-Austin) [mailto:wgselby@statesman.com]
Sent: Friday, September 25, 2015 2:41 PM
To: Lloyd Potter
Subject: FW: Fresh urgent reporter inquiry about 650 people moving to Texas daily
Dr. Potter:
I write afresh as we check a claim by the state comptroller, Glenn Hegar, that 650 people a day move to Texas.
His office says he drew this conclusion from data provided by Colorado-based IHS where an economist, Karl Kuykendall, told us by phone today that its experts simply presented information already put together by the U.S. Census Bureau based, Kuykendall said, on one-year American Community Survey results. He guided us to the migration data on the bureau site and we pulled down charts including one for 2013-14 suggesting there were 666 new residents of the state each day from July 2013 through June 2014 attributable to net migration both domestic/international.
I am writing you to see if this methodology delivers the most accurate estimate of the number of people lately moving to Texas daily. If not, what would be better and why? If so, are there wrinkles we should consider?
You will see some of the backup notes sent my way below this email.
Thanks again.
g.
W. Gardner Selby
Reporter / News
Austin American-Statesman
PolitiFact Texas
From: Lloyd Potter
Sent: Monday, September 28, 2015 12:38 PM
To: Selby, Gardner (CMG-Austin)
Subject: RE: Fresh urgent reporter inquiry about 650 people moving to Texas daily
Sorry for the delay.
I would steer toward the Census Bureau Population Estimates for what I think is a more accurate estimate of migration.
The most recent set of estimates (https://www.census.gov/popest/data/state/totals/2014/index.html) have about 239,104 net migrants (domestic and international) between 2013 and 2014. That would be about 655 net migrants (domestic and international) per day for the State. That’s what Karl pointed to with the PEPTCOMP search.
So 666 is not too far off. However, I couldn’t find from where the IHS numbers come. In searching through ACS migration data, it’s not too close to that.
Lloyd
Lloyd B. Potter, Ph.D.
Texas State Demographer
From: Selby, Gardner (CMG-Austin) [mailto:wgselby@statesman.com]
Sent: Monday, September 28, 2015 1:46 PM
To: Lloyd Potter
Subject: RE: Fresh urgent reporter inquiry about 650 people moving to Texas daily
Why do the figures fluctuate so?
Why is the CPS better for this than the one-year ACS?
In June 2014, we ended up with this story suggesting net migration of less than 500 people a day.
?
(Potter)
2:40 p.m.
CPS (Current Population Survey) asks more questions about migration behavior but probably does not provide a more accurate estimate of migration than ACS. The sample for CPS is about 50,000 annually and the Census Bureau does not tabulate state specific flows.
(http://www.census.gov/hhes/migration/data/cps.html). Usually, CPS estimates are for a calendar year (January 1 to December 31)
The ACS (American Community Survey) asks questions about where the respondent lived a year ago. The sample is larger than that of CPS but is still prone to errors associated with sampling. (http://www.census.gov/hhes/migration/data/acs.html). Usually estimates are for a calendar year (January 1 to December 31)
The population estimates produced by the Census Bureau are released with components of population change (net migration and natural increase). The estimates are generated using a number of different data sources (administrative records and survey data) and are generally thought to be and have been found to be reasonably accurate estimates of population and population change. (https://www.census.gov/popest/data/state/totals/2014/index.html). Census Bureau Population Estimates are usually for July 1 and population change (including migration) is estimated and reported from July 1 of one year to July 1 of the next.
So the migration component from the Census Bureau’s Population Estimates differs substantially in methodology from the ACS and the CPS and I would be surprised (perhaps pleasantly) if they were really close to each other.
If someone asked me to bet on which is likely to be the most accurate estimate of state migration, I’d put my money the Census Bureau’s Population Estimates.
From year to year, migration numbers do fluctuate, sometimes dramatically. Our economy was doing very well in 2013-14, bolstered by oil extraction, but was doing well even without. Next year, I’m anticipating migration will still be high, but we may see some muffling starting to show up with the reductions in extraction. Migration is usually and mostly driven by economic opportunity, thus, as our economy goes, we can expect migration to go.