Q&A of the Day – A Census Redo?

So the question isn’t about whether the previous census was accurate. It’s about whether the errors were intentional or unintentional.

Episode Transcript

Speaker 1 (00:03):
Have a question or topic you want to have addressed.
Just ask.

Speaker 2 (00:07):
This is the Brian Mud Show. Today's QNA. Since this redoo,
This is brought to you by I'm Listen as just
check mark collections. Each day I feature a listener question
that is sent by one of these methods. You may
email me Brian Mud at iHeartMedia dot com, hit me
up on social at Brian Mud Radio. May also use
the iHeart Radio talk back feature. Just go to w

(00:30):
JNO or via Patriot inside of the ihladio at we
love It, If you loved us when you get there
and make us your number one preset and then the
Brian Mudshow podcast number two preset. And while you're in there,
look for a little microphone button, see it tapping. You
may lay down the message right there, maybe for a
future QNA. Today's note this at Brian Mud Radio, please
explain deferential privacy with census and how that factored into

(00:55):
the read to outcome. Okay, so yeah, this is something
that came up that was actually cited by Florida's Attorney general.
I'll dive into that here momentarily. First, the latest on
President Trump's thinking about the census situation.

Speaker 3 (01:13):
The last census was held in twenty twenty. The next
one schedule for twenty thirty. President Trump says he might
want to speed it up, posting, I've instructed our Department
of Commerce to immediately begin work on a new and
highly accurate census. He continues, people who are in our
country illegally will not be counted in the census.

Speaker 2 (01:33):
As Fox's Lucas Tomlinson there and so as Othmeier said
in a letter recently to US Commerce, the US Commerce
Department calling for a correction to the twenty twenty census set.
In addition to skewing the data by utilizing dubious deferential
privacy and regularly delaying the publication of state population counts,

(01:56):
the twenty twenty national headcount and rodeously overcounted or undercounted
the populations of fourteen states. Florida was among those egregiously undercounted,
which deprives Florida of an additional House seat that its
citizens deserved to ensure fair representation in Congress and the
electoral College. Not only that, but this under enumeration also

(02:17):
costs Florida financially, given the census' role in allocating federal
funds to the States. Okay, so what is differential privacy?
How was it used in the twenty twenty census. While
deferential privacy is effectively a form of new math, it
was first formally introduced in two thousand and six by

(02:38):
FOURIDA different collaborators in a paper called Collaborating Noise to
Sensitivity and Private Data Analysis, and the concepts about as
easy as that title is too.

Speaker 1 (02:53):
The premise beyond the complicated concept is.

Speaker 2 (02:55):
To modernize data collection and analysis while protecting individual private
Used appropriately, the concept has several applications. It is regularly
used by major tech companies when collecting data.

Speaker 1 (03:07):
On users of its products.

Speaker 2 (03:09):
Is also regularly used in health studies to protect the
autonomy of patients. And it's defined like this as Joe
rubs his forehead. Deferential privacy is a mathematically rigorous framework
designed to release statistical information about data sets while ensuring
the privacy of individual data subjects.

Speaker 3 (03:31):
Oh my gosh.

Speaker 2 (03:32):
It guarantees that the inclusion or exclusion of any single
data point does not significantly impact the outcome of a
query or analysis, thereby protecting individual privacy within large data sets.

Speaker 1 (03:47):
Got all that, You're good? Oh? It's great? Yeah, The
head rubbing I thought would sync it in a little
more been replaced by a blind stare. It's noticeably.

Speaker 2 (04:01):
The new math had not been applied to census data
until the questionable twenty twenty census results. So here is
the not so insignificant thing about its application for census data.
These are known These are documented trade offs for using
this mathematical method.

Speaker 1 (04:22):
You are.

Speaker 2 (04:24):
Insinuating situating I should say, privacy over accuracy, privacy over accuracy.
When you're doing a census, what's going to be the
most important thing? Isn't accuracy going to be? I would think,
why are we so concerned with privacy anyway that I
so more noise or stronger privacy. Remember this is called

(04:48):
noise calibrating noise. More noise can reduce results accuracy. That
is a known trade off of this method. The complexity
designing private algorithms complicates the process that I think is understandable,

(05:11):
and the cost it has cost to the whole situation.
So as you can tell simply trying to explain what
the method was that was used to calculate the twenty
twenty census, that's complicated, not easily understood, right, And so
that certainly lends itself to potentially errors, intentionally or otherwise,
and it's notable that the first time this new method

(05:33):
was used for calculating census data, the results were different
than literally every expert's estimates heading into the census, As
was noted in a recent report by Florida Tax Watch.
They said the census undercount hurts Florida's political influence, demonstrating
that the twenty twenty census missed about seven hundred and
fifty thousand Floridians, or about three and a half percent

(05:56):
of the population. Correcting that error with the US Census
Bureau methodology shows the undercount shifted three US House seats Nationally. Colorado, Minnesota,
and Rhode Island would each lose a seed, while Florida, Tennessee,
and Texas would each gain one, raising Florida's delegation to
twenty nine seats instead of twenty eight. Every House seat

(06:20):
equals one electoral vote, so the miscount also diverted three
electoral votes away from Florida leaning states and the razor
a then twenty twenty four presidential contest that's wing alone
could have altered the electoral college result. Inside Congress, where
a landmark legislation has passed by a single vote, Florida's
missing representative could likewise decide national policy. So the question

(06:42):
isn't about whether the previous census was accurate. It's about
whether the errors were intentional or unintentional. But either way,
Floria's Attorney General of James Southmeyer, stated that it's time
for a correction. Othemayer suggested that it's possible and necessary
for the Conference Department to correct the previous census, perhaps
without conducting a new one. If, for example, the twenty

(07:03):
twenty census data still exists, it would be possible to
apply the traditional method of census calculations to determine a
more accurate outcome.

Speaker 1 (07:12):
Potentially right, but does it exist?

Speaker 2 (07:15):
So there's nothing in the Constitution that prohibits a correction
from taking place on the same data after the fact,
and it surely would be the most efficient way of
attempting to write the wrongs of the previous calculation. Now,
of course, Florida would stand to benefit most from a
newly conducted census that would take into account the post
pandemic population growth, which was the largest in the nation.

(07:37):
Though given the need for Congress to authorize a mid
decade census that remains extremely unlikely, so it's going to
be interesting to see what happens from here.