I don't think PUMS will work for me, given the lack of spatial detail.
I'm looking for evidence whether a lack of jobs/housing balance in
suburban areas hurts suburban employers. If it demonstrably does, then
this helps build the case for needing a mix of housing throughout the
region.
My hypothesis is that labor supply on the low-paying end is more
constrained in suburban areas that do not provide a mix of housing. If
there is a real labor supply constraint, this should show up as having
to pay a higher than average wage to low-paying occupations compared to
areas where jobs and housing are more balanced. If, on the other hand,
despite how it looks on a map, suburban employers actually have little
trouble attracting low-wage workers, regardless of how far they have to
travel (our low levels of congestion in the Kansas City area make this a
possibility), then there should be little or no wage differential.
My plan was to look at low-wage occupations and see whether retail
centers or corridors in more homogeneous suburbs paid higher wages than
retail centers/corridors in either urban areas or suburban areas with a
greater mix of incomes. Given that these retail centers do not cover
areas anywhere near as large as a PUMA, this plan clearly will not work
if the best I can do is use PUMS.
I think I can get away with using the earnings by industry table,
however, so long as I constrain my examination of wage differentials on
the low end of the wage scale to the retail trade industry.
Thanks to everyone who provided such great, informative and helpful
feedback!
Sincerely,
Frank
Frank Lenk
Director of Research Services
Mid-America Regional Council
600 Broadway, Suite 300
Kansas City, MO 64105
816-474-4240
www.marc.org
-----Original Message-----
From: Murakami, Elaine [mailto:Elaine.Murakami(a)fhwa.dot.gov]
Sent: Thursday, August 19, 2004 1:22 PM
To: ctpp-news(a)chrispy.net; Frank Lenk
Subject: RE: [CTPP] Worker Earnings by occupation
This is a situation where I think the best resource would be the Census
PUMS data. You won't get fine geographic detail, but you can group the
occupations and earnings to the classes that make sense for your region.
As you are part of the CB State Data Center, you should be able to get a
free copy of the PUMS DVD from the Census Bureau.
Elaine Murakami
FHWA Office of Planning
206-220-4460
-----Original Message-----
From: Peter H. Van Demark [mailto:peter(a)caliper.com]
Sent: Thursday, August 19, 2004 10:21 AM
To: ctpp-news(a)chrispy.net
Cc: Frank Lenk
Subject: Re: [CTPP] Worker Earnings by occupation
Frank:
>I am looking for the earnings of workers by occupation in the CTPP Part
>2 data, but cannot find it. Am I simply missing it, or is it really not
there?
Table P2011, Industry(15) by Worker Earnings in 1999 (12) comes the
closest. Only Tables P2003, P2009, P2015, P2016, and P2023 are by
occupation, and none are by earnings. There are twelve tables by
earnings, but none by occupation.
The TransCAD Table Chooser for CTPP Part 2 is a handy way to look at
tables by subject.
Peter
----------------------------------------------------------------
Peter Van Demark
Director of GIS Products and Training Phone: 617-527-4700
Caliper Corporation Fax: 617-527-5113
1172 Beacon Street E-mail: peter(a)caliper.com
Newton MA 02461-9926 Web site: http://www.caliper.com
_______________________________________________
ctpp-news mailing list
ctpp-news(a)chrispy.nethttp://www.chrispy.net/mailman/listinfo/ctpp-news
Another resource is the EEO data file prepared from 2000 Census data.
I just downloaded a sample file for our metro area with selected
occupations, by place of work, by earnings (distributional data), by
white/not hispanic and others. County and cities' data is only
available if population is 100,000 or more. But it is based on the full
Census sample which gives it an advantage over PUMS. A downside (for me
at least) is the amount of occupational detail, no summary groups
provided. But if you download to a spreadsheet you can sum up quickly
enough. The data tool makes it quick and easy to access the data. Data
tool:
http://www.census.gov/hhes/www/eeoindex/page_c.html More info:
http://www.census.gov/hhes/www/eeoindex.html
Regan Carlson
Sr. Research Analyst
Metropolitan Council/Research
(651) 602-1407
fax (651) 602-1674
>>> "Murakami, Elaine" <Elaine.Murakami(a)fhwa.dot.gov> 08/19/04 01:21PM
>>>
This is a situation where I think the best resource would be the Census
PUMS data. You won't get fine geographic detail, but you can group the
occupations and earnings to the classes that make sense for your region.
As you are part of the CB State Data Center, you should be able to get
a free copy of the PUMS DVD from the Census Bureau.
Elaine Murakami
FHWA Office of Planning
206-220-4460
-----Original Message-----
From: Peter H. Van Demark [mailto:peter(a)caliper.com]
Sent: Thursday, August 19, 2004 10:21 AM
To: ctpp-news(a)chrispy.net
Cc: Frank Lenk
Subject: Re: [CTPP] Worker Earnings by occupation
Frank:
>I am looking for the earnings of workers by occupation in the CTPP
Part 2
>data, but cannot find it. Am I simply missing it, or is it really not
there?
Table P2011, Industry(15) by Worker Earnings in 1999 (12) comes the
closest. Only Tables P2003, P2009, P2015, P2016, and P2023 are by
occupation, and none are by earnings. There are twelve tables by
earnings,
but none by occupation.
The TransCAD Table Chooser for CTPP Part 2 is a handy way to look at
tables
by subject.
Peter
----------------------------------------------------------------
Peter Van Demark
Director of GIS Products and Training Phone: 617-527-4700
Caliper Corporation Fax: 617-527-5113
1172 Beacon Street E-mail: peter(a)caliper.com
Newton MA 02461-9926 Web site: http://www.caliper.com
_______________________________________________
ctpp-news mailing list
ctpp-news(a)chrispy.nethttp://www.chrispy.net/mailman/listinfo/ctpp-news
_______________________________________________
ctpp-news mailing list
ctpp-news(a)chrispy.nethttp://www.chrispy.net/mailman/listinfo/ctpp-news
This is a situation where I think the best resource would be the Census PUMS data. You won't get fine geographic detail, but you can group the occupations and earnings to the classes that make sense for your region.
As you are part of the CB State Data Center, you should be able to get a free copy of the PUMS DVD from the Census Bureau.
Elaine Murakami
FHWA Office of Planning
206-220-4460
-----Original Message-----
From: Peter H. Van Demark [mailto:peter(a)caliper.com]
Sent: Thursday, August 19, 2004 10:21 AM
To: ctpp-news(a)chrispy.net
Cc: Frank Lenk
Subject: Re: [CTPP] Worker Earnings by occupation
Frank:
>I am looking for the earnings of workers by occupation in the CTPP Part 2
>data, but cannot find it. Am I simply missing it, or is it really not there?
Table P2011, Industry(15) by Worker Earnings in 1999 (12) comes the
closest. Only Tables P2003, P2009, P2015, P2016, and P2023 are by
occupation, and none are by earnings. There are twelve tables by earnings,
but none by occupation.
The TransCAD Table Chooser for CTPP Part 2 is a handy way to look at tables
by subject.
Peter
----------------------------------------------------------------
Peter Van Demark
Director of GIS Products and Training Phone: 617-527-4700
Caliper Corporation Fax: 617-527-5113
1172 Beacon Street E-mail: peter(a)caliper.com
Newton MA 02461-9926 Web site: http://www.caliper.com
_______________________________________________
ctpp-news mailing list
ctpp-news(a)chrispy.nethttp://www.chrispy.net/mailman/listinfo/ctpp-news
Frank:
>I am looking for the earnings of workers by occupation in the CTPP Part 2
>data, but cannot find it. Am I simply missing it, or is it really not there?
Table P2011, Industry(15) by Worker Earnings in 1999 (12) comes the
closest. Only Tables P2003, P2009, P2015, P2016, and P2023 are by
occupation, and none are by earnings. There are twelve tables by earnings,
but none by occupation.
The TransCAD Table Chooser for CTPP Part 2 is a handy way to look at tables
by subject.
Peter
----------------------------------------------------------------
Peter Van Demark
Director of GIS Products and Training Phone: 617-527-4700
Caliper Corporation Fax: 617-527-5113
1172 Beacon Street E-mail: peter(a)caliper.com
Newton MA 02461-9926 Web site: http://www.caliper.com
I am trying to see whether one part of our region pays a wage premium
because it imports workers from longer distances. Typically, to compare
wages, one looks at average earnings by occupation. I am looking for
the earnings of workers by occupation in the CTPP Part 2 data, but
cannot find it. Am I simply missing it, or is it really not there?
Sincerely,
Frank Lenk
Director of Research Services
Mid-America Regional Council
600 Broadway, Suite 300
Kansas City, MO 64105
816-474-4240
www.marc.org
Jiji:
Ken is totally correct. You need to convert the CTPP commuter matrix
into something like an "observed home-based work" trip table using trip
rates per employed residents from your local travel survey, the NCHRP
#365 (1.50 trips/worker), or the NHTS. Remember that the census only
provides a count of workers, and is not a count of work trips, per se.
We (SF Bay Area) did *not* use the CTPP2000 Part #3 commuter matrices
in our Year 2000 model validation project. Our model validation project
ended in April 2004, and the CTPP Part 3 data was available May 2004.
Such is life. On the other hand, we did use the 5% PUMS data (available
August 2003) to construct what we needed: county-to-county commuters by
means of transportation by household income quartile. We then converted
these commuter matrices into "observed home-based work" trips. For our
validation, we sequentially validated our auto ownership (to Census PUMS
data), work trip generation, work trip distribution, and work trip mode
choice models to these census-based observed databases. We then went
through some fairly exhaustive "network validation" stages where we
compared the model simulated traffic (daily and peak) and transit
(daily) to observed traffic and transit counts. At this network
assignment validation stage we diverged from our "observed HBW" trips in
order to better match observed traffic/transit databases. The basic
problem we were having with the PUMS-based observed work trips was that
it was too low (transit, especially) in our Transbay corridor (SF
to/from the East Bay). This is a problem with either too few commuters
in the decennial census, or perhaps our trip rate for this distance of
work trip was too low.
So, my recommendation is to use the commuter matrices from the CTPP,
convert them into appropriate "observed home-based work" trips, then do
initial stages of validation against these observed databases, BUT be
prepared to diverge from these CTPP-based observed trip tables if and
when your transit/traffic assignment validation is inadequate. OR, you
can always "hold true" to the CTPP-based observed data, and focus your
attention on calibration / hammering away at your non-work trip models.
Cheers,
Chuck Purvis, MTC-Oakland
>>> Ken Cervenka <kcervenka(a)nctcog.org> 08/16/04 10:46AM >>>
On your question #1: One thing to keep in mind relates to the fact
that a
"HBW trip" for modeling purposes represents a trip end, e.g., a Home to
Work
trip and a Work to Home trip would actually be counted as two separate
trips, i.e., HBW productions and two HBW attractions.
So, using your NCHRP example, 1,000 employees would generate 1,500 HBW
trips--but only half of these (750) would be comparable to a
CTPP-derived
home-to-work trip.
In terms of why travel survey data never supports 2.0 HBW trips per
employee: It is partly because of absenteeism on an "average"
weekday,
partly because some jobs are simply not five-day-a-week jobs, and
partly
because some (or even many) people don't actually go straight from home
to
work (or from work to home) on their surveyed day (e.g., they engage in
some
trip chaining, which turns some Home-to-Work trips into HNW and NHB
trips,
and turns some Work-to-Home trips into NHB and HNW trips).
Hope this helps!
sincerely,
Ken Cervenka
NCTCOG
-----Original Message-----
From: ctpp-news-bounces(a)chrispy.net
[mailto:ctpp-news-bounces(a)chrispy.net]On Behalf Of Jiji Kottommannil
Sent: Monday, August 16, 2004 12:09 PM
To: ctpp-news(a)chrispy.net
Subject: [CTPP] CTPP Part 3
Hi All,
I am a new user of the CTPP data and am in the process of comparing
CTPP
part 3 tract-tract worker flow data to the Home-Based Work trip table
from the travel demand model I am currently working on. I came across
a
few questions/issues for which I would welcome your
thoughts/experiences:
1) Do the CTPP part 3 flows account for all the home-based work trips?
I
noticed that the commonly accepted home-based work attraction rate is
about 1.5 trips per total employment (NCHRP 365) and this creates more
HBW trips than there are in the CTPP part 3. Let me know if I there is
something wrong in my interpretation of the data.
2) How much success have the MPO's had in calibrating their trip
distribution models based on CTPP Part 3? What level of accuracy is
typically desired?
Thanks!
Jiji
Jiji V. Kottommannil
Transportation Modeling Specialist,E.I.T.
Crawford Bunte Brammeier
Phone: 314-878-6644, Ext. 38
Fax: 314-878-5876
Elaine, Joe,
Although it was not an in-depth view and analysis, I did undertake a
quick check of employment here in the Minneapolis-St. Paul area.... more
to see what was coming out of the CTPP files as compared to what we
received from the state Employment Securities offices. Since we utilize
the state figures and NOT the CTPP numbers, the differences have not
been an issue. I can't offer much of an explanation other than to
relate our own findings. In any case they did NOT seem to be random.
Two checks were made. The first one was entirely with the 2000
figures. I mapped the Employment Securities numbers by TAZ and then the
CTPP part 2 numbers by TAZ. In a fairly sizeable number of instances
the two files were substantially different. However, the differences
seem to have been off-set by combining contiguous TAZs. That is to say,
one zone was under-counted by 1500, but the adjacent was over-counted by
1800. Admittedly I simply chalked up the bulk of these as being
"address matching problems". Correctly or not, that was the assumption
I made. Nearly all of these major differences fell into this
situation.
The second check was using the 1990 CTPP and 2000 CTPP figures.
Sorting the set "high to low" by differences surprised me at first. The
greatest gain was 13,319 employees; the next 6 TAZs showed gains from
4,000 to 5,600. Upon closer inspection they mostly proved to by TAZs
where major growth actually did occur. The TAZ with the 13,000 gain was
around the airport where our CTPP generated employment numbers always
seemed suspect. We never use them. Why they are so far off the beam I
don't know. Airports seem to be touchy, quirky activity centers.
Those TAZs at the other end of the spectrum (showing major declines)
appeared to be located adjacent to TAZs with major gains. Again, I
chalked it up to address matching. Maybe I'm being lazy, for it could
be something else! In several instances there were TAZ boundary issues
that played into the mix. In a few other TAZs, there actually were
major declines. For example, a rather large publishing company
relocated from the inner city to a suburban location. That move did
cause a big drop in employment that did NOT rebound.
What came out of this review was that in the overwhelming majority of
instances, mapping TAZs with major drops to those with major increases
resulted in a pattern. They seemed to be contiguous to each other. For
those that weren't, there were determining causes that could be
identified. The TAZs were random around the region, but they affected
TAZs with employment and were usually balanced out with adjacent TAZs.
Maybe our region was lucky. You might want to check to see if this
same pattern checks out with adjacent TAZs, if you haven't already done
so. In the case of New York City, I can't imagine there being much of
an address matching problem.
Bob
Average Vehicle occupancy is lowest for HB-Work trip purpose.
You need to use household travel survey data for calculating AVO for other trip purposes. You should use the NYMTC household travel survey data from the late 1990s, and compare with 2001 NHTS results (see page 11 of the "Highlights of the 2001 NHTS" report http://nhts.ornl.gov/2001/reports.shtml
You can run the 2001 NHTS data for the NY/NJ Metro area alone, using the on-line table generator on the same web-site listed above. Use the "sub-group" option and select HHC_MSA= 5602
Elaine Murakami
FHWA Office of Planning
206-220-4460
-----Original Message-----
From: Daniel Evans [mailto:devans(a)dot.state.ny.us]
Sent: Wednesday, August 18, 2004 9:03 AM
To: ctpp-news(a)chrispy.net
Subject: Re: [CTPP] CTPP Part 2 data for New York City
Does anyone know if Average Vehicle Occupancy (AVO) rates, not merely
Journey to Work, can be derived from the CTPP?
Thanks,
Dan
Daniel Evans
Senior Transportation Analyst
Mid-Hudson South Planning Coordinator
New York Metropolitan Transportation Council
(voice) 718-472-3198, (fax) 718-482-7431
_______________________________________________
ctpp-news mailing list
ctpp-news(a)chrispy.nethttp://www.chrispy.net/mailman/listinfo/ctpp-news
Does anyone know if Average Vehicle Occupancy (AVO) rates, not merely
Journey to Work, can be derived from the CTPP?
Thanks,
Dan
Daniel Evans
Senior Transportation Analyst
Mid-Hudson South Planning Coordinator
New York Metropolitan Transportation Council
(voice) 718-472-3198, (fax) 718-482-7431
