********** C E N S U S 2 0 0 0 B U L L E T I N
**********
Vol. 2 - No. 45 Nov. 24,
1998
Citing a need to reduce the workload for "the large
household followup" and expected overall coverage
improvements, especially among hard-to-enumerate
populations, the Census Bureau has decided to adopt a
six-person questionnaire design. This will apply to both the
short and long forms in the Census 2000 plan, which includes
scientific sampling, and in an alternative plan, which calls
for traditional census-taking methods.
The Census Bureau anticipates that the change from a
five-person to a six-person questionnaire for forms that are
mailed out or delivered by enumerators to housing units for
mailback will cut followup workload for large households in
half. Planning estimates put the number of mailback
households with seven or more persons at slightly more than
1 million households versus about 4 million households with
six or more.
The issue was revisited recently during discussions about
ways that the Census Bureau might increase coverage in a
census that does not include statistical sampling to
supplement traditional methods. With deadlines for
advertising printing specifications scheduled for October
1998, it was found to be more cost-effective to require
six-person forms only, notwithstanding a final decision on
sampling. The alternative would have necessitated printing
two sets of questionnaires for the entire country, a
prohibitive expense.
Other advantages of the six-person questionnaire:
--it retains the design initiatives developed in the
commercially designed form to make an easy-to-complete,
respondent-friendly questionnaire.
--it can be introduced into the Census Bureau's basic system
for data capture and mailback questionnaire processing
without major disruption.
--it provides for a slight positive advantage in the
Integrated Coverage Measurement survey one of the components
in the current sampling plan through more timely data
capture of complete large households.
--it reduces respondent burden by requiring six-person
households to respond only once by using a mailback
questionnaire designed for households with six persons,
rather than one for five (which would have required
additional reporting by the six-person households).
Disadvantages of the six-person mailback questionnaire are
few. Although followup is reduced by an estimated 50
percent, there are overall higher costs associated with the
six-person mailback questionnaire due to the long form data
capture method, which involves capture of each page
regardless of the number of persons in the household.
Also, the additional width of the short-form is another area
of higher costs. Both, however, were considered to be
relatively minor cost increases when balanced against data
quality erosion and losses resulting from the two-stage
enumeration for large households.
For further information about Census 2000 Bulletins, contact
J. Paul Wyatt of the Public Information Office on
301-457-3052 (fax: 301-457-3670; e-mail: pwyatt(a)census.gov).
Statewide TAZs. TIGER/Line can only take ONE definition of a TAZ, so we
thought that the metropolitan TAZs were the most important to include in
TIGER/Line. So, definition of state TAZ is a tabulation question for CTPP
that will be handled by Phil Salopek in Journey to Work, and not the Census
Bureau GEOGRAPHY division.
We are trying to get an estimate for how many areas will WANT to define state
TAZs--this question is part of the letter that went out through the FHWA
Division office. We hope that you have included this information in your
response.
thanks for your input!
1. when?--we are hoping that the software and the TIGER/Line 98 CDs will ship
at the end of January. MPOs and State DOTs will have six months to return the
files to the Census Bureau.
2. Minimum threshholds? We are still waiting to hear from the Census Bureau
Disclosure Review Board. When I went to APDU, there was some discussion that
Block Group recommendation was to have threshhold of about 600 persons.
However, Phil Salopek of Census Bureau Journey to Work thinks that CTPP may
not have to meet the same requirements.
Ed:
Here in Denver we have had a few problems with county boundaries because
Denver is a consolidated city/county. Back when they could annex, they
created weird boundaries. We have one area where the county line goes
through an apartment building! Therefore, we have a few zones that cross
county lines. We have dealt with this in the past by creating partial
zones in each county and will plan on doing that again for 2000.
Larry Mugler
-----Original Message-----
From: ed c [SMTP:berwyned(a)mcs.com]
Sent: Monday, November 23, 1998 4:03 PM
To: ctpp maillist
Subject: [CTPP] defining TAZs
As many of you know we (MPOs) will soon be in the throws of
defining our Traffic Analysis Zones (TAZs) within our urban
areas. When working through that process one constraint,
that is new for 2000, will be that TAZs respect county
boundaries. That is, a TAZ can not cross a county
boundary. Is this something that you can live with? Will
this be a big problem for any regions?
Here in northeastern Illinois we have tended to cross over
many different political boundaries over the years but we do
respect county boundaries in our planning work.
thanks
ed christopher
please respond to the list (all recipients) so others may
benefit from any comments.
Hi folks,
If perhaps you're interested in the opinion of someone who knows more about
census geography/data than about TAZs specifically:
I think TAZs should nest to census tracts (which by definition will then
follow county lines). Tracts need more than one TAZ where there are
employment concentrations; more than one tract may constitute a TAZ when
you're in homogenous residential areas within a municipality.
I also think you should be worrying, or at least thinking, about the
sampling errors involved in small population TAZs. Block group data, even
with an average of 600 housing units, are too small to deliver decent data.
This is especially true for the journey to and place of work data, which
have additional non-response errors beyond the overall sampling and
non-response errors for the long form. I have never reported or relied on
census-based place of work counts unless there were at least 1000 workers
in the geographic area being reported.
My guess is that census will end up permitting data to be delivered on a
CTPP/UTPP for whatever geography you draw; disclosure will probably not be
a problem unless you start using the finely drawn race/ethnic groups which
will now be possible. But just because you can get the data doesn't mean
the numbers are worth anything.
One final word--at the risk of incurring the wrath of at least some of
you--I really don't see why it's so important to maintain complete
comparability from one set of TAZs to the next. The most important thing,
I would think, is to have TAZs which reflect the current land
use/population distribution/etc. to make good forecasts. If they need
changing for whatever reason, they should be changed.
Patty Becker
=============================
Patricia C. (Patty) Becker 313/535-2077
APB Associates/SEMCC FAX 313/535-3556
17321 Telegraph #204 Home 248/355-2428
Detroit, MI 48219 pbecker(a)umich.edu
Those of us who were part of the Alpha Test of the TAZ-UP (TAZ definition
software) are now reviewing the first draft of the User Guide. One item that
I am trying to draft right now is about why a TAZ for CTPP may be different
than a TAZ that you have for your regional model.
This is my current draft, which I am sure will change, with your input. Also,
IMPORTANT NOTE--we still have not heard from the Census Bureau Disclosure
Review Board what kind of population threshholds or "average" populations they
may require for tabulation.
Oh, and finally, the plan is for the TAZ submissions to Census Bureau to occur
around June/July of 1999. Then the TAZs will be added to TIGER/Line 99, which
will be distributed for you around early 2000 to validate that what you
submitted was correctly input to TIGER.
-------------------------------
Geographic units
Standard Census products use standard Census geography. That is, tabulations
are prepared by State, county, munipality, census tract, and block group.
The CTPP allows users to define their own geography. In this case, traffic
analysis zones (TAZ). A TAZ defined for CTPP may differ from a TAZ defined
for a local model.
Reasons why this may occur:
1. A TAZ defined for CTPP 2000 is required to be composed of the smallest
allowable polygons in the Census TIGER/Line file. Typically, in the
TIGER/Line file, block boundaries are limited to physical features such as
streets, power lines, streams and rivers, etc. TIGER/Line does NOT have parcel
boundaries. A local GIS may have more detailed geography, e.g. parcels and
lot lines that allow TAZs to be defined by unique parcels. You will want to
try to make your TAZs for CTPP 2000 as close to your own TAZs as possible.
Issues: You may want to include businesses on both sides of a street. You
cannot create another TIGER line segment to break up the smallest allowable
polygon to capture both sides of the street. You are limited to selecting the
entire smallest allowable polygons on both sides of the street segment.
DRAW A PICTURE HERE.
2. Small TAZs with little or no resident population or employment
population. For your local model, you may have defined TAZs where you expect
future growth to occur. This makes it better for you to display your
forecasted data to 2010 or 2020. However, for the tabulation of Census
responses in year 2000, for areas with very small populations, there is a risk
that the data will be suppressed by the Census Bureau. You may want to
aggregate areas to create somewhat larger TAZs for CTPP so that you can take
advantage of the large sample (1 in 6 households nationwide) from the
decennial long form. That is, you will get data reported in the CTPP for a
larger geographic area, and will have to disaggregate it to your *real* TAZs
later on.
Similarly, if ALL your TAZs are defined with very small populations, you risk
the chance that likewise, ALL your data will be suppressed by the Census
Bureau. For example, your local model has 300 zones, but the average
population or employment is less than 300 persons. You run a high risk of
having all your data suppressed due to confidentiality requirements of the
Census. For the CTPP, you may want to have a *super-district* system that
reduces the number of zones to 150 zones. With your best guess, this
increases the average population and employment to 500 persons. This would
likely not be suppressed by the Census Bureau.
Ed:
Counties are not a problem for our TAZ layout for the Cincinnati metro
planning area. Our boundaries were configured using the counties (8 of
them).
Census boundary compliance will be the bigger issue as our TAZs were
initially drawn in the 60s and the census boundaries have changed more
in the mean time. However, we have begun a project to adapt our TAZs
to the 2000 census areas. I am coordinating the Census Statistical Areas
Review in 6 of our counties and am on the committees for the other 2.
With this knowledge of tract and block group boundaries, I expect that we
can revise most TAZ boundaries to match.
I anticipate our main problem will be the 'special' zones that we have set
up for commercial and industrial areas that follow land use edges instead
of physical features. This caused problems with the 90 CTPP where I
had to pick a TAZ for blocks that were split by TAZ boundaries. I did this
by giving more importance to the employment component of the block as
our use for the CTPP was place of work data. Thus many of these
'special' zones had households that should have been in the adjacent
zones. I accounted for this in preparing my demographic tables, but
others using our CTPP wouldn't have a clue about this problem.
By the way, I E-mailed my TAZ Update participation form to Tom Mank at
FHWA as his name and address were on the last page. Let me know if it
should go elsewhere.
Don Burrell
OH-KY-IN Regional Council of Govmts
Cincinnati
Agree with Chuck. County boundaries are relatively permanent, unlike
muncipal limits. The last time County boundaries were adjusted in
California was in the late 1940's, I believe, and even then the changes
were minor. A county was dissolved about 100 years ago. This seems
stable enough for travel forecasting work.
/Steve Colman
===
"If you find something you like, buy a lifetime supply-- they're going
to stop making it." --Bob Dolfay
As many of you know we (MPOs) will soon be in the throws of
defining our Traffic Analysis Zones (TAZs) within our urban
areas. When working through that process one constraint,
that is new for 2000, will be that TAZs respect county
boundaries. That is, a TAZ can not cross a county
boundary. Is this something that you can live with? Will
this be a big problem for any regions?
Here in northeastern Illinois we have tended to cross over
many different political boundaries over the years but we do
respect county boundaries in our planning work.
thanks
ed christopher
please respond to the list (all recipients) so others may
benefit from any comments.
If you are planning on attenting this year's Transportation
Reseach Board meetings be sure to support the activitities
of the data committees. Here is a link to all of the
information that you will need to know. Looking forward to
seeing you at session 133. Feel free to distribute this
link those you know that might be interested.
http://www.mcs.com/~berwyned/census/trbpgm99.html
(note this is from the preliminary program)
ed christopher
