2nd Annual ESRI Federal User Group Meeting
Call for Presentations
April 29-30, 1999
National Geographic Society, Washington, D.C.
YOU MAKE IT HAPPEN!
An important component of a user group meeting is the exchange of
information among users. Many years of GIS experience are collectively
shared and disseminated through presentations for users by other users.
The 1999 ESRI Federal User Group Meeting is a unique opportunity for you
to meet with other ESRI federal users to discuss important issues and
exchange information on how GIS technology can help solve complex problems
within your industry.
We would like to invite you to share your GIS project or application. If
you are interested in giving a presentation or demonstration of your
federal GIS project or application, please send us the information listed
below via e-mail to dpaxson(a)esri.com <mailto:dpaxson@esri.com> or FAX to
909-307-3074 to Dana Paxson, no later than Friday, February 5, 1999.
Name:
Organization:
Phone:
FAX:
E-mail:
Presentation Description:
More details on the meeting will follow in the next month. We look forward
to seeing you in Washington, D.C.!
Todd Rogers
Federal Business Development
<mailto:Trogers@esri.com> Trogers(a)esri.com
2070 Chain Bridge Road, Suite 180
Vienna, VA 22182-2536
Tel 703-506-9515 Fax 703-506-9514
Nathan and the CTPP newsgroup:
A few comments on Nathan's 12/9/98 posting.
First, we hear what you are saying, Nathan, regarding your interest in
NYSDOT in getting block group data, over and above what the MPOs receive.
We are trying to see how we might accomodate this; we'll keep everyone
posted through this newsgroup on the progress.
Secondly, however, I have to point out that a number of the assumptions you
have regarding the production of the CTPP and its relation to STF3 are just
plain wrong. The CTPP is --NOT--repeat NOT-- part of STF3. The CTPP is
programmed by a different group of people, at a different time, independently
of anything done for STF3. The only thing they have in common is that they
both use the same source file, our internal sample edited detail file.
We formatted the CTPP to look like STF3 to make it easier for people who
were accustomed to using Bureau summary tape files. But the whole idea
behind the CTPP is that it is CUSTOM (read non-standard) content for
CUSTOM geography. The content is tailored to transportation planning uses
and consists of much more detailed cross-tabulations than the single
variable distributions found on STF3.
Even if the Bureau produced a workplace-based STF3 as you suggest (an idea
which has been rejected in the past), my understanding is that the content
would still be lacking for transportation planning uses. So while this is
a legitimate issue for people interested in place of work data generally,
it's not really a CTPP issue.
On the other hand, although we make the CTPP content available for
standard Census geographic units like tracts or block groups, most MPOs
choose to have the data created for TAZs, a non-standard or custom
geographic unit. So even if the content of STF3 was more detailed (which
for 2000, it will not be) the geography wouldn't be as useful for
transportation planners and there would still be a need for a CTPP. So it
seems to me that there's really little point in talking about STF3 in
relation to CTPP. They aren't related.
Similarly, there is no standard census product that provides
county-to-county work flows. The file produced in December of 1992 that
Chuck mentioned (STF-S-5) was again a special tabulation done out of this
office for a number of customers, one of which was BEA whom I believe
produces the REIS CD you mentioned. (In other words, if we don't produce
the special tab, there is no data for them to put on their CD.)
This product lacks the means of transportation, travel time, and peak/off
peak information provided in CTPP Parts 3 and C, which I assume
transportation planners generally find to be valuable. This product was
produced by the CTPP programmers (again having nothing to do with STF3) and
would have been done considerably later, IF AT ALL, had it relied on
non-CTPP staff. Yes, it probably delayed the first releases of CTPP by a
few months, but as Chuck indicated, it was a product that was used by the
transportation planning community as well as others like BEA, BLS, etc.
For 2000, I assume that as in the past, there will not be a county commuter
flow product created as a standard product. If sponsors are found for such
a product, then the likelihood that it will be created is increased. My own
opinion is that if such a tabulation is going to be produced, we should try
to broaden its scope and include characteristics of the commuters between
counties, not just counts. I think this would make the product more
marketable to a larger audience. However, it would also likely increase the
size and complexity of the dataset.
We might also consider producing commuting flows for lower-level geography
(say, places above a certain size), as part of such a product, but I haven't
really seen the demand for this. I'd be inclined to say that what we do in
the statewide CTPP for place-level flows is sufficient.
You also mention that Census geographic units should always be available as
a summary level. Well, this is your opinion and you're entitled to it, BUT,
I think it's a rather narrow view. First, I think the evidence indicates
that this is a minority view in the transportation planning community. In
1980 we produced something called Part V of the UTPP which was data by block
group of work, even where the rest of the package was tabulated for TAZs. I
know of very few people who used these data. In 1990 we did something
similar, producing Part 7 of the CTPP in all urban packages which was data
by tract of work. Use of Part 7 has been isolated at best. So it is clear
to me that TAZs are of primary importance and there is little demand for
standard census geography summaries as you suggest.
I have comments on some of the other points you made, but this is too long
already, so I'll end here. But I encourage you and others to continue to
voice your sentiments and concerns. We may not always agree, nor be able to
implement your suggestions, but we will listen and try to make improvements.
--Phil
an archive version of the ctpp news is now available at
http://www.chrispy.net/ctpp-news/
it is not glamorous but it does the job. for those like me
that do not store book marks very well, you can find the
link on our web site on the page that talks about the mail
list. here is the direct link to that page
http://www.mcs.com/~berwyned/census/maillist.html
Date: Mon, 14 Dec 1998
From: Census2000 <Census2000(a)ccmc.org>
House Leaders Consider Shifting Census Oversight to Internal
Watchdog Committee
Monitoring Board, House Panel Continue Hearings
House Republican leaders are considering a proposal to
transfer jurisdiction over the census to the House Oversight
Committee, which is responsible for congressional office
budgets, campaign finance reform and contested elections,
and operations of the House generally. The pending change
was reported last week by the Capitol Hill newspaper Roll
Call and confirmed by lawmakers and staff aides who work on
census issues. Under the proposal, the Subcommittee on the
Census of the Government Reform and Oversight Committee,
created a year ago, would be eliminated.
Rep. Carolyn Maloney (D-NY), the senior Democrat on the
census subcommittee, strongly criticized the proposed
shift. "The Leadership felt their control on this issue
slipping away. People are becoming more aware of their ploy
to be sure America's minorities remain anonymous," Rep.
Maloney said in a written statement. She accused Republican
leaders of "burying the Census issue in a paper pushing
committee." Census Subcommittee Chairman Dan Miller (R-FL)
has not yet issued a statement. He and Rep. Maloney were
presiding over a hearing in Miami, FL, last Thursday to
examine local efforts in support of the census.
The House Oversight Committee has six Republican and three
Democratic members; the Government Reform panel's current
party ratio is 24 to 19, plus one Independent. Chaired by
Rep. William M. Thomas (R-CA), the Oversight panel has no
subcommittees. The committee's senior Democrat in the 105th
Congress, Rep. Sam Gejdenson (D-CT), is leaving the panel.
Roll Call also reports that Republican leaders might change
the panel's name back to the Committee on House
Administration, the title used when Democrats were in the
majority.
Responsibility for overseeing Census Bureau activities in
the House of Representatives has shifted several times in
recent years. Prior to 1995, the Subcommittee on Census and
Population (later called the Subcommittee on Census,
Statistics, and Postal Personnel) of the Committee on Post
Office and Civil Service had jurisdiction over the Census
Bureau and broader issues related to the Federal statistical
system. The entire committee was eliminated when
Republicans assumed the majority in 1995, and jurisdiction
over the census was handed to the Subcommittee on National
Security, International Affairs, and Criminal Justice of a
revamped Government Reform panel. The national security
subcommittee was chaired by former-Rep. William Zeliff
(R-NH) in the 104th Congress, and then by Rep. J. Dennis
Hastert (R-IL) for the first year of the 105th Congress.
Republican leaders created the new census subcommittee at
the start of 1998 and named Rep. Miller as chairman. It is
unclear whether the House Oversight Committee would monitor
all Census Bureau activities and other data issues or just
the 2000 census. Another subcommittee under the Government
Reform panel has been examining changes to the Federal rules
governing the collection of racial and ethnic data.
Monitoring Board continues hearings: The Census Monitoring
Board will hold a hearing in Sacramento, CA, on December 16,
to discuss the nearly-completed census dress rehearsal.
Sacramento was one of three sites where the Census Bureau
conducted a dry-run of census operations this year. Under a
congressional directive to prepare for two census designs,
the Bureau carried out its original plan in Sacramento,
which included sampling to complete follow-up visits to
unresponsive households and a post-census quality check
survey to measure and correct under-and overcounts. The
Board's hearing will take place from 8:00 a.m. - 12:30 p.m.,
in the City of Sacramento Council Chambers, 915 "I" Street.
The meeting is open to the public.
Congressional hearings continue: The House Subcommittee on
the Census held a hearing in Miami, FL, last Thursday to
discuss ideas from local officials and community
organizations on how to ensure an accurate census.
According to a December 11 article in The Miami Herald,
witnesses from community-based organizations and Rep. Carrie
Meek (D-FL) told the lawmakers that statistical methods must
be used in addition to the traditional mail and door-to-door
counting. Rep. Maloney praised a letter in support of the
Census Bureau's 2000 plan signed by nearly two-dozen local
groups including the Urban League, the Alliance for Aging,
the Haitian American Foundation, and the Miami-Dade County
Hispanic Advisory Board. Chairman Miller said that more
aggressive promotion could improve the accuracy of the
census and noted that two Federal district courts concluded
this summer that sampling was unlawful.
At a press conference earlier in the week, Miami-Dade County
Mayor Alex Penelas announced the formation of a Complete
Count Committee to oversee and coordinate local activities
in support of the 2000 census. Mayor Penelas said the
area's net undercount of 70,000 in 1990 "shortchanged" the
community. "For a successful Census that includes everyone,
we need to use statistical techniques such as sampling," he
noted. The Mayor was joined by representatives of community
organizations including the NAACP, American Association of
University Women, Organization of Chinese Americans, and
Cuban American National Council, Inc. All of the groups
stressed the importance of census participation in their
communities and voiced support for the use of sampling to
supplement the direct counting effort.
The subcommittee plans additional field hearings to discuss
ideas from local officials and community-based organizations
on improving the census process. At the December 3 - 4
meeting of the 2000 Census Advisory Committee, Chairman
Miller's staff announced sessions in Phoenix, AZ, on January
28, and Los Angeles, CA, several days later. The hearings
were disclosed before word of the subcommittee's possible
elimination became public.
Congressional committee assignments: Republican and
Democratic lawmakers continued to make committee assignments
in anticipation of the start of the 106th Congress on
January 6. Three new Senators will join the Committee on
Governmental Affairs, which oversees the census: Sen. Judd
Gregg (R-NH), who also chairs the appropriations
subcommittee responsible for the Census Bureau's budget, and
freshmen Senators George Voinovich (R-OH) and John Edwards
(D-NC). The new committee members will replace Senators Sam
Brownback (R-KS) and Don Nickles (R-OK) and retiring Sen.
John Glenn (D-OH), the panel's senior Democrat.
The Census 2000 Initiative will publish a roster of all
congressional committees with responsibility for census
funding and oversight after the 106th Congress convenes and
panel assignments are complete.
Questions about the information contained in this News Alert
may be directed to TerriAnn Lowenthal at (202) 484-2270 or,
by e-mail at <terriann2k(a)aol.com>. Please direct all
requests to receive News Alerts, and all changes in
address/phone/fax/e-mail, to Census 2000 at
<Census2000(a)ccmc.org> or 202/326-8700. Please feel free to
circulate this information to colleagues and other
interested individuals.
if you have ever looked at the table headings they are sometimes very
difficult to figure out what they actually contain. For instance, if you
want the JTW number of workers by mode there really isn't a table
heading for this. JTW is part C and table C1 is time to work. Its heading
is "Time leaving home to go to work-total and peak period by means of
transportation to work"...universe workers 16+ who did not work at
home. Each cell contains the number of workers, not the time they spent
going to work by that mode; the table coordinates tell you that the table
contains 2 tables sequentially first coordinates 1,1 to 1,19 contain the
total and then the peak period with table coordinates 2,1 to 2,19.
Table headings should be immediately intuitive!!!!!
What needs to be done is that the table headings for C1 must say what
the table contains:" workers, leaving home to work by mode of travel;
daily and peak period" universe of workers 16+ who did not work at
home. This way you immediately know the contents are workers all day
and peak period
Get an English teacher to write or review the headings, the products
produced by Census are written in Census Jargon which at best is
unclear because it tries to be overly precise and interpretation can be
very tough even knowing what you are reading.
>>> ed c <berwyned(a)mcs.com> 12/10/98 06:20pm >>>
below is a dialogue that has been going on that i believe would
be of interest to the entire ctpp mail list. it is extremely
topical because just today Elaine Murakami called to remind me
that i need to get something written up regarding our survey of
several trb committees regarding the content (tables) of the ctpp
package. here is a web page that presents some preliminary work
in that area. don did remind me of an early product that even we
in chicago used (stf s-5).
http://www.mcs.com/~berwyned/census/notes/content.html
needless to say any comments folks have on content for the ctpp
(the tables) would be appreciated. i also recommend that you post
your comments to the list
ed christopher
------------------------------------------------
Don,
Let me see what I can answer right now and what has to go into
the hopper:
> Place of work outside SMSA of residence. This needs a further
> breakdown. Our Cincinnati area is a 13 co. CMSA so the data for
> "outside each pmsa" doesn't tell me if the worker is in the
other
> pmsa. Similarly, we are adjacent to the Dayton msa and I
can't
> tell if the workers are coming to or from there. The county and
> major city breakouts in this table are also valuable. (CBD?)
> These concerns apply mainly to the standard tabulations. I
> strongly recommend that the STF-S-5 commuter tabulations
> be prepared again for nation-wide county to county work
> trip commuting. I actually depend on
> these for my regional in and out commuting and, consequently,
> the regional net employment.
I think the work flows have to come from the county to county
file, and I would assume that will be repeated. It will make
sure it's on our list. Actually, I was never aware of it as a
formally named product (STF5?) but I did have a copy of it for
Michigan counties. The issue of grouping counties as areas on
the CTPP has to be fought out within the transportation planning
community.
> Another issue is continuation of the non-motorized travel modes
> of biking and walking along with working at home. (Assuming
> that transit and carpool modes will be continued.) These are
> becoming more important in our region as we deal with air
> quality conformity, congestion reduction and sprawl.
Bicycle, Walked, and Worked at home are on the census dress
rehearsal questionnaire, which means that unless something very
unusual and unlikely happens, they will be on the 2000 long form.
Once the data are collected, it's really up to the transportation
planning community to decide how they will be tabulated on the
CTPP. Also, you are going to have more opportunity to do your
own tabulations via the Internet, so if you need
something special you will probably be able to do it (albeit for
a fee), or pay someone to do it for you.
> Finally, an issue that I am assuming will be fixed with the
> 2000 CTPP is place of work coverage in the suburban
> and rural areas of our metro area. In 1990, less than half
> of a county's workers were 'assigned' to a TAZ of work
> in three of the 8 counties in our MSA.
With nationwide TIGER coverage, I believe that all workplaces
with street addresses are to be coded to tract/block. This of
course permits their assignment to TAZ. The question is what
happens with the inadequate workplace address entries on the
census. You might want to talk to Phil Salopek about their plans
on this. In 1980 Phil Fulton devised an elaborate allocation
algorithm which fixed a lot of the problems in these data for
workplaces inside the urbanized area, and with the full TIGER
coverage the algorithm can be run for everywhere. However, if
you have some major workplaces in the more rural counties that
don't have addresses and that you want to be sure are coded
correctly, you may be able to provide some workplace name
information to the Bureau to assist
with this.
> And that brings to mind another final issue. In 1990, our CMSA
> and MPO included the same 8 counties. In 1993, five more
> counties were added to the CMSA, but not to our MPO
> planning area. Therefore, our 2000 CTPP
> will not include the added CMSA counties. Does this matter?
The CMSA and the MPO don't have anything to do with each other,
really. The CMSA is defined under OMB rules (now under review
for possible significant changes; a federal register notice is
due very soon); there can be a considerable amount of politics in
the MSA/PMSA/CMSA designations because a lot of people think "the
bigger the better" which makes for larger CMSAs. If you're not
planning for the additional 5 counties, you probably don't need
the CTPP for them. Besides, won't some other MPO have it (if the
5 counties are their own PMSA, they probably have some MPO
coverage somewhere).
> Thanks for asking!
Hope this helps!
> (copied to Ed Christopher)
Ed, if you want to post this to the list, please feel free.
=============================
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
below is a dialogue that has been going on that i believe would
be of interest to the entire ctpp mail list. it is extremely
topical because just today Elaine Murakami called to remind me
that i need to get something written up regarding our survey of
several trb committees regarding the content (tables) of the ctpp
package. here is a web page that presents some preliminary work
in that area. don did remind me of an early product that even we
in chicago used (stf s-5).
http://www.mcs.com/~berwyned/census/notes/content.html
needless to say any comments folks have on content for the ctpp
(the tables) would be appreciated. i also recommend that you post
your comments to the list
ed christopher
------------------------------------------------
Don,
Let me see what I can answer right now and what has to go into
the hopper:
> Place of work outside SMSA of residence. This needs a further
> breakdown. Our Cincinnati area is a 13 co. CMSA so the data for
> "outside each pmsa" doesn't tell me if the worker is in the
other
> pmsa. Similarly, we are adjacent to the Dayton msa and I
can't
> tell if the workers are coming to or from there. The county and
> major city breakouts in this table are also valuable. (CBD?)
> These concerns apply mainly to the standard tabulations. I
> strongly recommend that the STF-S-5 commuter tabulations
> be prepared again for nation-wide county to county work
> trip commuting. I actually depend on
> these for my regional in and out commuting and, consequently,
> the regional net employment.
I think the work flows have to come from the county to county
file, and I would assume that will be repeated. It will make
sure it's on our list. Actually, I was never aware of it as a
formally named product (STF5?) but I did have a copy of it for
Michigan counties. The issue of grouping counties as areas on
the CTPP has to be fought out within the transportation planning
community.
> Another issue is continuation of the non-motorized travel modes
> of biking and walking along with working at home. (Assuming
> that transit and carpool modes will be continued.) These are
> becoming more important in our region as we deal with air
> quality conformity, congestion reduction and sprawl.
Bicycle, Walked, and Worked at home are on the census dress
rehearsal questionnaire, which means that unless something very
unusual and unlikely happens, they will be on the 2000 long form.
Once the data are collected, it's really up to the transportation
planning community to decide how they will be tabulated on the
CTPP. Also, you are going to have more opportunity to do your
own tabulations via the Internet, so if you need
something special you will probably be able to do it (albeit for
a fee), or pay someone to do it for you.
> Finally, an issue that I am assuming will be fixed with the
> 2000 CTPP is place of work coverage in the suburban
> and rural areas of our metro area. In 1990, less than half
> of a county's workers were 'assigned' to a TAZ of work
> in three of the 8 counties in our MSA.
With nationwide TIGER coverage, I believe that all workplaces
with street addresses are to be coded to tract/block. This of
course permits their assignment to TAZ. The question is what
happens with the inadequate workplace address entries on the
census. You might want to talk to Phil Salopek about their plans
on this. In 1980 Phil Fulton devised an elaborate allocation
algorithm which fixed a lot of the problems in these data for
workplaces inside the urbanized area, and with the full TIGER
coverage the algorithm can be run for everywhere. However, if
you have some major workplaces in the more rural counties that
don't have addresses and that you want to be sure are coded
correctly, you may be able to provide some workplace name
information to the Bureau to assist
with this.
> And that brings to mind another final issue. In 1990, our CMSA
> and MPO included the same 8 counties. In 1993, five more
> counties were added to the CMSA, but not to our MPO
> planning area. Therefore, our 2000 CTPP
> will not include the added CMSA counties. Does this matter?
The CMSA and the MPO don't have anything to do with each other,
really. The CMSA is defined under OMB rules (now under review
for possible significant changes; a federal register notice is
due very soon); there can be a considerable amount of politics in
the MSA/PMSA/CMSA designations because a lot of people think "the
bigger the better" which makes for larger CMSAs. If you're not
planning for the additional 5 counties, you probably don't need
the CTPP for them. Besides, won't some other MPO have it (if the
5 counties are their own PMSA, they probably have some MPO
coverage somewhere).
> Thanks for asking!
Hope this helps!
> (copied to Ed Christopher)
Ed, if you want to post this to the list, please feel free.
=============================
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
Greetings Ed Christopher:
Please place me on the CTPP e-mail list server so I can contribute to it
and view others correspondence.
Thanks,
Bob Frey
Massachusetts Highway Department
bob.frey(a)state.ma.us
At Ed Christopher's suggestion I am posting this to the list server.
The following are some observations on the E-mail traffic I have been
observing with respect to the CTPP summarization issues that I have
raised with Ed at and since his sub-committee meeting at TRB last
January. I must note that more recent e-mail is now starting to mention
non-TAZ needs
Phil Salopek in a recent e-mail says if you want Census geography
instead of TAZ you don't have to participate in the TAZ up
program....well at NYSDOT we want the statewide element of the CTPP
to be detailed to CDP or lower and all the urban area elements for our
use to be detailed to Census geography....tract and block group as the
lowest level in the urban area.
NYSDOT is interested in Census geography at the block group level
especially in urban areas. Since each of the MPOs will be responding
separately to the CTPP TAZ questionnaire, in addition to NYSDOT, then
MPOs will indicate their need for TAZs or census geography. This way
the MPOs can get what they want and the state can get what it needs.
The glitch is if Census or Phil Salopek says the state has to have it the
MPO way.
Frankly I don't see the problem....the CTPP is part of the STF3. As part
of STF3 it is summarized for CTPP parts A and 1 to the block / block
group level and everything above from the get go. All that is required is
the production of STF3 for the work place. County work flow is
produced very early on in Census processing prior to the CTPP. These
data are historically available on the REIS CD. So for the most part
Census provides CTPP part C or 3 at the county level outside of CTPP.
Clearly they have the routines in place to do it by specifying lower
geography with selected focus (such as state or metro area). So its not
just a CTPP processing software issue.
The real issue is getting Census to recognize several things:
1 Census geography should always be available as a summarization
level for all locations
2 If you are not an urbanized area 50,000+ then you likely do not have
an MPO, and you do not have TAZs even though you may be a small
urban area
3 TAZs are unique to "MPOs with models" and they don't define TAZs
them the same way
4 If you are not now an urban area now and eventually become one in
say 2002 then you are out of luck if TAZ is the only summarization level.
This may be especially painful when we move to the ACS.
5 If you need to add TAZs , combine or subdivide them later you can't
6 This is not just an MPO product
7 This is really a summarization level issue like congressional boundaries
8 The problem that Census really needs to address is how to aggregate
data up from lower geography for values such as mean and median;
where summarization is really not possible ( median income or median
travel time for example). Resolving this problem then enables Census or
the end user to create the data at the lowest possible geographic unit
and provide a summary level routine for computational values. All the
MPOs need then is a simple lookup table to relate TAZs to Census
geography and a routine to fix the computational values
9 From a purely economic product perspective Census geography
means you can sell the product to third parties and recover the cost.
TAZ is a summarization level unique to MPOs. There are lots of value
added resellers of STF3 data now in addition to Census. The CTPP
creates the workplace version of STF3 for a number of measures that
are important to not only transportation planners but also to anyone
interested in who, how and/or what arrives at work place locations.
10 Remember to make sure that the CTPP coding uses Census FIPS
codes and not Census Place codes that Caliper used in 1990, otherwise
it will not directly match common tiger based coverages.
11 The CTPP software that extracts the data table components from the
CD rom needs to be improved on the data extraction part. Often one
wants some range of Fips or town county, state codes for either the
residence , work place or JTW and for selected fields (e.g. total trips,
transit). As the 1990 CTPP software works you get all geography that
you can point to on the map, not bad on the statewide side except if you
want trips from selected origins outside the state. If you are using the
urban element then selection is much more difficult because of the level
of detail. In either case JTW from selected to selected geography is not
easily do-able except in a database product. Some accommodation on
the direct data selection side rather than through the visual graphical
user interface would be helpful. Something like "extract" the Census
routine for use with the Census STF CDs would be helpful. Otherwise
you really need a database product like SAS, Paradox, dBASE or
Access to get at the data the way you want.
12 What would really help would be Census creating STF3 for the
workplace and the present county work flow data for lower
summarization levels
13 Lastly it might be useful to note that at NYSDOT many of the
information requests I receive come from the economic development
interests. These requests reflect transportation or government concerns
for attracting businesses into NYS. They come from within our agency or
from our State Economic Development agency. These questions typically
ask about workers first and then where they live. NYS can't be that
different from the rest of the states that compete for the location of new
jobs. Many of these requests are for areas just outside of the urban
boundary. The fact that these type of informational questions are being
asked illustrates the value of detailed Census geography and
summarization levels beyond what has been provided for at the
metropolitan planning through TAZs.
As always I am available to discuss some of the more mechanical
technical aspects of using the CTPP data in non TAZ analyses with
anyone who wants to discuss why one size does not fit all.
The following message is from Thabet Zakaria <DVRPC>, not from Margaret
Shaw. Please direct all correspondence regarding this mailing to Mr.
Zakaria, NOT to Ms. Shaw. Mr. Zakaria's e-mail address is:
tzakaria(a)dvrpc.org
Thank you.
Development of TAZs for the 2000 CTPP
This is a brief description of the Traffic Analysis Zones (TAZ) in the
Delaware Valley Region, which includes an area of 3,833 square miles and a
population of about 5.2 million. I have prepared this summary in the hope
that other areas in the country will find it useful for preparing a zonal
system for the 2000 CTPP.
In 1960, the Penn Jersey Transportation Study developed a set of TAZs for
data collection and travel simulation within the cordon area (urbanized
area) which did not consider county boundaries. The basic unit was a
modified grid square, 0.10 mile on a side in city centers and 0.50-mile
elsewhere. A total of 162 districts (TAZs) were aggregated to 51
Superdistricts for evaluation and presentation.
In the late 1970s, DVRPC realized that it could not continue to collect
massive data for transportation planning, and must depend on secondary
sources, including the Bureau of the Census. It was decided to adopt
census geography and convert the old TAZs to census tracts and block
groups. Census tracts were considered adequate for regional travel
simulation, except for the Philadelphia CBD, where block groups were used
to define TAZs. The cordoned area was expanded to the region's county
boundary and divided into 1,330 TAZs, which were aggregated to 71 county
planning areas to simplify analysis and presentation. This system of TAZs
was expanded in the 1980s, because the Bureau of the Census in cooperation
with the counties divided some census tracts for the 1990 Census. This
system proved to be easier to work with and facilitated the preparation of
a correspondence table for the 1990 CTPP between TAZs and blocks.
It should be noted that the new TAZ system is not free of problems. Some
tracts are too large or too small for network coding and modeling, and they
must be reviewed each decade because of tract splitting. Also, some tracts
are too narrow or curvilinear, and are thus not suitable for travel
simulation purposes. It is essential that each tract in the area be
examined regarding its layout, size, type and magnitude of development.
Several criteria could be used to develop TAZs form census geography, such
as population, employment, trip production, trip attraction, zone centroid
location, and trip loading (network grain). DVRPC used an average of about
4,000 persons and 3,000 employees per zone in the development of its TAZ
system. Obviously, the size of TAZs should be smaller for corridor and
sub-area studies and small MPOs.
