I want to echo the sentiment expressed by Mr. Moore in Pueblo. Whatever size we set TAZs at, and however we use our GIS to create "subTAZs" and fill them with PL-94 and private data, I do not want to see "N/A" on any of them. I would also like to have some idea of what the threshold is where we might get synthetic data instead of real estimation. I for one would rather set my geography bigger where necessary to get actual data. I suspect that, using local knowledge, I can do a better job of estimation than somebody in an office a long way from my area generating "synthetic data." Imputation has its uses, but I would rather know what is real and what is made up.

Here at Metroplan we are currently planning to work at two levels. We will create "Data TAZs" at whatever population/employment level the BTS, FHWA, etc. set for accurate figures. We will then use PL-94 and private data to fill our "Model TAZs," which will be smaller but enclosed within the Data TAZs.

My summary point: I will accept whatever standard is set for TAZ geography, but please give me a clear idea of what that standard is. If there needs to be a trade-off between TAZ size and accuracy, it would help if I could be given some guidelines, like "TAZs below 1,200 population will include a large share of imputation," or "TAZ's 1,500 and above will be done with little or no imputation," etc.

That said, synthetic data is a lot better than no data, and if I have to accept imputations to get a 2015 CTPP, I will go with it. Then again, why am I worrying about the years after 2015 when I've got worries enough right now?

Jonathan Lupton
Metroplan
Little Rock AR
501-372-3300


Ed Christopher wrote:
Bill--VERY WELL PUT!!!  I hope others read this and heed your words.

Bill--this will not solve your dilemma but there is serious research
work going on and we certainly need more to make it a reality but there
is a likely prospect that the data for a 5-year product will be
synthetic.  This would allow for very teeny tiny TAZs that would be "DRB
free".  I suppose that is probably a little better than our zones being
"data free".  Kidding aside, one of the big reasons the push back on the
proposed DRB rules for our 3-year data is so important is because even
through the 3-year geography is way bigger (20K) we need some
potentially real data at some workable level that is reasonably useful. 

Bill Moore wrote:
  
Ed:

I probably wasn't very clear in explaining what I see as the potentially serious issue for small/medium MPO areas that are using (or developing or planning to develop) travel demand models.  We currently have 306 TAZs ranging in size from a couple of square blocks in the CBD to (literally) a few hundred square miles encompassing entire census tracts in the outermost reaches of the county.  On average, the HH/TAZ is around 200 (~450 population/TAZ).  As in most places, there is a lot of variation arouind that mean.

What I'm concerned about is "seeing" a map of our existing TAZs with dozens - if not hundreds - of them filled in with N/As, based on the minimum DRB criteria which were being discussed.  For many of them, I am concerned that even with the 5-year ACS accumulations they will not rise to the DRB minimums.  And, this comes at a time when both private developers and public officials are demanding even FINER granularity (than the 306) for TD model use and/or forecasts of socio-economic variables.

In some cases, we are being asked to "draw" future TAZs where there are now only miles of grazing land and only vague generalities about the quantity or quality of future land uses, target markets, etc.  AND, of course, to "populate" those future TAZs (replete with all the necessary demographics for modeling ~70,000 new households) for about the next 30-100 years.  If I have to climb out on that limb, I'd prefer to have some rationale (like travel behavior and demographics in current TAZs that might resemble the future ones) for SWAGging at the problem (for 47 square miles and a guesstimated minimum of 1.1 million trips/day).

'Tis a puzzlement, indeed - especially without the prospect of more finely grained local data from the 2010 Census and CTPP as a beginning point.  There are two other MPOs - one larger and one smaller - that I know of who are facing similar dilemmas in their planning areas.  We'll all be following this closely.

Bill

-----Original Message-----
From: ctpp-news-bounces@chrispy.net
[mailto:ctpp-news-bounces@chrispy.net]On Behalf Of Ed Christopher
Sent: Thursday, June 12, 2008 3:14 PM
To: ctpp-news@chrispy.net
Subject: Re: [CTPP] CTPP Update

Bill--I am sorry if I was not clear.  What is being discussed are three
size TAZ
Large-65K
Medium-20K
Small- the size we all grew up with

The large sizes would be for 1 year ACS data that is only released for
areas of 65K or more. The medium size TAZs would be for 3-year ACS data
which is only being released for areas of 20K or more and our Small TAZs
would be for 5 year ACS data which the Census Bureau will release for
tracts and block groups.

As far as CTPP data products we are looking at something similar to what
we used to get every 10 years but using 5-year ACS data and a similar
product using 3-year ACS data.  right now we are in the throws of having
our 3-year data tables reviewed by the Census Disclosure Review Board
which is turning out to be a prelude for the 5-year product.

I hope this clarifies things but if not let's please keep this dialogue
going because it is important for all us to understand.

Bill Moore wrote:
    
Ed:

>From a small (~135,000 pop) MPO perspective, the implications of these super-TAZs for travel demand modeling are devastating.  If 65K, we would have two TAZs; if 20K, we would have six or seven.  Fortunately, we are also an attainment area and are not mandated to do modeling at all; however a new TMA with population of 200,000 would be so mandated and would have a total of three or 10, respectively, so it would become virtually impossible to do modeling for AQ conformity.

The four MPOs (howdy, Larry) along the Front Range are trying to develop an integrated model for a population of ~3.5 million, so at 65K would have only 54 TAZs, or at 20K only 175.  The latter number is far less than we now have (306) in the Pueblo MPO - the smallest of the four.

Bill Moore, MPO Administrator
PACOG MPO/TPR

-----Original Message-----
From: ctpp-news-bounces@chrispy.net
[mailto:ctpp-news-bounces@chrispy.net]On Behalf Of Ed Christopher
Sent: Wednesday, June 11, 2008 3:33 PM
To: ctpp-news@chrispy.net
Subject: Re: [CTPP] CTPP Update

Larry--Yes, when we creat 20,000 and 65,000 TAZs our goal is to get
tablulations for these areas.  Or course the DRB would have to approve
any tables.  The current problem with the ACS 1 and 3-year data (due out
at the end of the year) that the Census releases is that your data
coverage in like swiss cheese.  You get data for some areas and not
others.  Our plan is to have complete coverage to eliminate this problem
with any tables paid for with CTPP dollars. Who knows maybe we can get
the Census Bureau to see the light and adopt the concept of having zonal
coverage consistent with the data release thresholds for their data
products as well.

Larry Mugler wrote:
      
If we create "super" tazs with populations of 20,000 and/or 65,000,
would we be able to get tabulations for such areas?

Larry G. Mugler, AICP - Planning Services Manager
303-480-6759 - Customer Resource and Support
Denver Regional Council of Governments

-----Original Message-----
From: ctpp-news-bounces@chrispy.net
[mailto:ctpp-news-bounces@chrispy.net] On Behalf Of Murakami, Elaine
Sent: Wednesday, June 11, 2008 12:43 PM
To: ctpp-news@chrispy.net
Subject: RE: [CTPP] CTPP Update

Re:  poverty and Income comparisons between Census 2000 Long Form and
ACS:
The only report I am familiar with is:
http://www.census.gov/acs/www/Downloads/Report05.pdf
Pages 34-37 discuss county-level comparisons.  These are comparisons
between the Census 2000 Long Form (aka sample) and the ACS "C2SS", and
18 of the test ACS counties.  They found that the C2SS results
consistently had lower incomes than the Census 2000 sample, and
therefore, a higher poverty ratio.   They found that income allocation
occurs 30 percent in the Census 2000 sample, and 24% in the C2SS.

I do not know if the Census Bureau has pursued other research on how ACS
respondents answer the income questions.

Elaine

-----Original Message-----
From: ctpp-news-bounces@chrispy.net
[mailto:ctpp-news-bounces@chrispy.net] On Behalf Of Ed Christopher
Sent: Wednesday, June 11, 2008 9:50 AM
To: ctpp-news maillist
Subject: [CTPP] CTPP Update

On May 21 and June 4 members of the CTPP team conducted a series of
webinars to help bring our community up-to-speed on several of the CTPP
activities.  The slides from these presentations can be downloaded from
our webinar room at http://fhwa.acrobat.com/ctpp1 (enter as a guest) and
the responses to many of the questions asked during the meetings are
presented below.  If folks have additional questions please feel free to
post them to the list serve and someone on the CTPP team will attempt to
answer them.

CTPP Webinar - Questions and Answers

Question: Will the 3-year CTPP data product based on ACS include tables
by race and household income?
Response:  Yes but the best source for information on race will still be
the Census Bureau's standard ACS tabulations found on American
FactFinder (http://factfinder.census.gov/).  The 3-year CTPP proposed
tables for parts 1 and 2 include a few tables with the variable
"minority status." Also, there are various tables with the variables -
Income, Poverty Status and Worker Earnings that will be available.
Please note that Poverty Status is a calculated variable based on
household income, family size, number of children and number of family
members over the age of 65.

Question: Since the new CTPP products will be based on data collected
over 3 years (or 5 years), to which year will the income be adjusted?
Response:  Income will be adjusted to the last year of the survey
period.   For the first 3-year product all incomes will adjusted to
2007.

Question: Is there data available on zero car households? If yes, what
geography level is it available?
Response:  Yes, there will be information available for zero car
households in both the 3-year CTPP data product and as part of the
standard Census products.  The data will be available for geographic
areas greater than 20,000 residents.

Question: Slides 6 & 7 of the DRB presentation show different
percentages of data lost due to DRB rules, why is that?
Response:  Slide 6 showed the number of workers lost while slide 7
showed the number of origin-destination pairs lost.  The point is that
when thresholds were applied to CTPP Part 3 data as was the case with
the 2000 data, many folks looked at the loss of workers as being
significant but more surprising was actual loss of individual O-D
pairs.  All in all table thresholds devastated the part 3 flow data.

Question: 2005 ACS data doesn't have information on Group Quarters, but
the 2006 and 2007 do.  How is this being handled in the 3-year 2007 ACS
products?
Response:  The Group Quarter data available for 2 years (2006 and 2007)
will be reweighed to account for the missing year.

Question: Will the variable - Means of Transportation to Work have walk
and bike modes combined? Also, what about taxi, bus and other?
Response:  Regarding the 3-year CTPP data product there are multiple
category lists for the variable - Means of Transportation to Work.
There are a few tables all 17 modes plus the total are shown separately
but then there are many others where walk and bike have been collapsed
together.  A great deal of the detail on the mode variable is in the
hands of the Census Bureau's disclosure Review Board.  Currently they
have proposed some very tight restrictions on the proposed 3-year CTPP
tables.  Please see http://trbcensus.com/drb/ for more

Question: Do you have anything on the TAZ definition timeframe that you
can share with the group?
Response:  Please see our Status Report newsletter for the latest
http://www.TRBcensus.com/newsltr/sr052008.pdf.

Question: Can you talk about the difference in income between ACS and
Long Form?
Response:  In Census 2000, the question was asked during April for the
previous year, 1999.  Because April is so close to IRS annual return
data it was felt that "good" income data was being collected because it
was fresh in people's minds.  However, for the ACS the same questions
are asked, but the respondent could be getting the survey during any
month of year (depending on when she/he received the form).  This
reported income is then adjusted to the current year (year of
tabulation) based on CPI.  There are several issues with this with the
largest being that at anytime during the year most people do not how
much money they made in the last 12 months.  Needless to say the income
question is messy.

Question: Poverty rates seem discontinuous, higher. Any ideas why?
Response:  Yes, they are discontinuous.  Hopefully they will look better
with the 3-year ACS data products.

Question: If the DRB is so strict with its rules then why bother with
new TAZs?
Response:  We hope that the DRB will relax its rules but having TAZs
that correspond to the ACS data release products (65,000 pop TAZs,
20,000 pop TAZs and small TAZs) will help us in the future all sorts of
data products.  Having geography match the data release seems to make
good sense so that we can at least get complete coverage (wall-to-wall)
within a region.

Question: What are the implications of rising fuel prices on travel
choices on mode to work?  Has anybody thought about it, especially with
aggregating 3 years worth of data?  Since the ACS asks about "usual"
mode, it could still miss modes used only part of the time.
Response:  The 3 year ACS trend data might look a lot different than the
2000 data. This might be a good research question.

Question: Is the category list for the variable - occupation being
consolidated?
Response:  No, DRB has not asked us to collapse the variables - industry
or occupation.

Question: Is anybody planning to write to Congress about the DRB issue?
Response: AASHTO SCOP is sending a letter to the Census Bureau to appeal
the DRB decision.  As FHWA staff, we cannot contact Congress directly
but we know that many regional agencies do talk with their Congressional
delegations.  It is somewhat ironic but many of the congressionally
mandated analysis like the FTA New Starts program and environmental
justice analysis all need data at a smaller geography level which will
be difficult to obtain with the DRB's current rules.
Question: We are relatively a new MPO and I have a CTPP 101 question.
As an MPO do we need to provide you any information, and what are our
sources of information regarding CTPP specifically regarding TAZs?
Response: The CTPP listserve is a good place for information regarding
ongoing CTPP related activities. The TAZ definition process is planned
for March/April of 2009 and it is recommended that MPO's allocate some
budget when preparing their work plans for FY 2009, especially for new
MPOs. The CTPP Status Report is also a good source of information.  If
you have any particular questions please do not hesitate to contact
anyone one of us on the CTPP team.

Question:  Somewhere it was noted the first CTPP ACS product would be
available in 2009 for areas of 20,000 or more for places and counties.
Response: Yes that is correct.  We are still in the negotiation stage
regarding some DRB issues but this is our GOAL.  This first CTPP product
would use 3 years of ACS
data.

Question:  There was also a mention of TAZ level data also for 20,000
population threshold.  However, I am not sure if this is included in the
first 3 year product for 2009, or whether the first TAZ level report
will not be available until 2012 with a 5 year product.
Response: To get "small" area geography like a TAZ or a census tract,
the census requires 5 years of ACS data. So, we are planning to wait for
2006 thru 2010 ACS data to incorporate 2010 Census geography and
weights, which we hope would be tabulated by 2012.   This will likely
require some data synthesis before release.  Because of the probability
that the data will be synthetic, we do not anticipate any population
threshold.

--
Ed Christopher
Resource Center Planning Team
Federal Highway Administration
19900 Governors Drive
Olympia Fields, Illinois  60461
708-283-3534 (V)  708-574-8131 (cell)
708-283-3501 (F)

_______________________________________________
ctpp-news mailing list
ctpp-news@chrispy.net
http://www.chrispy.net/mailman/listinfo/ctpp-news

_______________________________________________
ctpp-news mailing list
ctpp-news@chrispy.net
http://www.chrispy.net/mailman/listinfo/ctpp-news

_______________________________________________
ctpp-news mailing list
ctpp-news@chrispy.net
http://www.chrispy.net/mailman/listinfo/ctpp-news
        
--
Ed Christopher
Resource Center Planning Team
Federal Highway Administration
19900 Governors Drive
Olympia Fields, Illinois  60461
708-283-3534 (V)  708-574-8131 (cell)
708-283-3501 (F)
_______________________________________________
ctpp-news mailing list
ctpp-news@chrispy.net
http://www.chrispy.net/mailman/listinfo/ctpp-news

This e-mail transmission (including any attachments) contains information that is confidential and may be legally privileged. It is intended for the use of the addressee only. If you received this e-mail in error, we request that you contact us immediately by telephone or return e-mail, and that you delete this message from your computer. If you are not the intended recipient, please be advised that any dissemination, distribution, or copying of this e-mail is strictly prohibited.

_______________________________________________
ctpp-news mailing list
ctpp-news@chrispy.net
http://www.chrispy.net/mailman/listinfo/ctpp-news
      
--
Ed Christopher
708-283-3534 (V)
708-574-8131 (cell)

FHWA RC-TST-PLN
19900 Governors Dr
Olympia Fields, IL 60461

_______________________________________________
ctpp-news mailing list
ctpp-news@chrispy.net
http://www.chrispy.net/mailman/listinfo/ctpp-news

This e-mail transmission (including any attachments) contains information that is confidential and may be legally privileged. It is intended for the use of the addressee only. If you received this e-mail in error, we request that you contact us immediately by telephone or return e-mail, and that you delete this message from your computer. If you are not the intended recipient, please be advised that any dissemination, distribution, or copying of this e-mail is strictly prohibited.

_______________________________________________
ctpp-news mailing list
ctpp-news@chrispy.net
http://www.chrispy.net/mailman/listinfo/ctpp-news