Dear T,
I completely agree with your comments on the serious limitations
of ACS data. Comparing to your desire to apply the ACS data for
transportation, I am very disappointed about the quality and limitations
of ACS to serve as reference data for evaluating our annual county and
municipal estimates of population and households. However, there are
many other data users out there hoping to get timely demographic
characteristics for making decisions involving either to spend money or
to get money. Less accurate statistics information is still better than
no information for decision making.
Sample statistics from the decennial census long form (from 1/6 of
households or 17 million household samples in Census 2000) indeed are
very much better and reliable (accurate) than that from the 838,000
samples from the 2004 ACS survey in a single year. Without the 2004 ACS
data, we can only depend on the Census 2000 sample statistics. Do you
like that? That is your choice. If so, just don't bother about the ACS
survey data for transportation applications. I am still hope to find
something worthwhile from the annual ACS survey data to check on the
annual estimates of population and households. It is still better than
the March Supplement CPS (Current Population Survey) data which only
contain about 800 (household) samples for Colorado. As of 7-1-2004
Colorado has more than 2,026,000 households. Comparing to CPS, I
appreciate very much to have ACS.
How about canceling the ACS program and campaigning to re-install
Long Form for Census 2010? You all know that if the Senates do not fund
or under fund the ACS test programs for year 2006 and the years after,
it is highly likely that the Census Bureau has to abandon ACS program
and request funding to work on Long Form for 2010 Census. How many of
you would like to see that happens?
>>> <TMarchwinski(a)njtransit.com> 9/17/2005 12:51 PM >>>
Dear Mr. Lin- I disagree with you about census long form vs. ACS. I
cannot use the ACS for any transportation planning analysis except maybe
to see where general trends are going. Even there, looking at NJ data
by county, there are some weird changes in mode data even from year to
year. I use the 2000 Census data almost every week, and the ACS, unless
it can get down to a smaller geography, does not do much good. Unless
you are doing some high level comparisons about average travel time or
average mode split etc., it cannot be used for what most transportation
planners need, such as trip distribution, mode split, auto occupancy,
vehicle ownership etc. If there is no long form, then we will have to
take what we can from the ACS and combine it with other data, surveys
etc.
-----Original Message-----
From: ctpp-news-bounces(a)chrispy.net
[mailto:ctpp-news-bounces@chrispy.net] On Behalf Of Richard Lin
Sent: Friday, September 16, 2005 7:48 PM
To: ctpp-news(a)chrispy.net; tzakaria(a)dvrpc.org;
Elaine.Murakami(a)fhwa.dot.gov
Subject: RE: [CTPP] 2004 ACS Data Release
Dear All,
We all realize the limitations and quality of the ACS data. Also,
we all are seeking better resolutions. First of all, we hope the Census
Bureau gets enough funding to carrying out ACS operations and to
improving the quality of data by accumulating greater samples. Hope by
Census 2010, the quality of the ACS data is equal to or even better than
that of the traditional long form. Without annual ACS survey, we will
fall back again to the once a 10-year long form data availability. Will
you prefer still using the Census 2000 Long Form data in 2005 or do you
appreciate the limited 2000-2004 ACS data? Without the annual ACS data,
we will go back to the dark again.
The good news is that if the Census Bureau gets funding to
continue ACS survey through Census 2010 then we should have much more
reliable and larger sample ACS data for better quality time series.
With the sampling error variation, we should be more cautious about
time series fluctuations and verify the true changes in numbers such as
the Jew Jersey's Mercer County having lesser workers riding bikes or
walk to work (5,450 or 3.5 in Census 2000 comparing to 2,924 or 1.9 in
ACS 2004). We should also be cautious when we compare apple to orange.
We may check the data at state (NJ) level by comparing the state's
Census 2000 numbers and ACS 2000 numbers to establish a ratio factor for
county adjustment for time series comparison.
>>> "Murakami, Elaine" <Elaine.Murakami(a)fhwa.dot.gov> 9/14/2005
1:39:12
PM >>>
Dear Zak,
Thank you for providing your opinion on the disutility of the ACS 2004
data for your applications, and I am glad that others are examining data
for their area and are sharing their findings with the listserv.
My earlier email was sent to CAUTION people who might try to compare
2000 decennial data to ACS 2004 data. As I said in my earlier email, it
is probably better to compare ACS results to ACS results, than to
compare ACS results to decennial census results, however, results from
ACS from 2000 are very limited. The geographic coverage of County level
data from the ACS in 2000 is sparse, which makes comparisons for ACS
between 2000 and 2004 possible only in some areas.
The ACS did not go into "full implementation" until 2005, and it will
require multiple years of data collection before data for small areas
will be reliable because of the small samples collected each month and
averaged over time. As Tom Marchwinski pointed out, averaging over
multiple years creates other problems. In 2005, the CB begin
differential non-response follow-up, so that in areas (tracts) with low
mail-back responses, there is a greater rate (1:2 and 2:5 instead of
1:3) made to follow-up non-respondents from the mail-back and CATI
portion of the ACS. Not surprisingly, the areas with low mail-back
responses are more likely to be low income, and higher shares of African
American and Hispanic populations. My guess is that this will cause a
shift in numbers between 2004 and 2005, and could impact variables such
as carpooling and transit use and number of vehicles in households.
The Census Bureau recommends that the ACS should be used to describe
characteristics, and not to use it for COUNTS. When I examine the
Mercer County NJ data comparing Census 2000 (workers in households) to
ACS 2004 (workers in hhlds), I find that about the only thing to say is
that "driving alone" appears to increase from 77 to 78%, and
"carpooling" to decrease from 11 to 10%, "rail" also appears to increase
from 4 to nearly 6%. (Worked at Home is not included in column in the
table below). HOWEVER, this is not taking into account the effects of
the different survey methods, where, generally speaking, the decennial
census has a greater share reporting "carpooling," which is why a
BRIDGE from decennial 2000 data is so important. Thus, it is probably
incorrect to say that carpooling is declining.
Mercer County, NJ
Workers in Hhlds Census 2000 ACS 2004
Number Pct Number Pct
Total 153,665 153,041
Drove alone 118,390 77.0 119,597 78.1
2-person CP 13,105 8.5 12,026 7.9
3+ person CP 4,580 3.0 2,722 1.8
Bus/trolley bus 4,585 3.0 4,390
2.9
Streetcar/trolleycar+ 195 0.1 1,168 0.8
Railroad or ferry 6,105 4.0 8,795
5.7
Bike or Walk* 5,450 3.5 2,924 1.9
Taxi/motorcycle/other 1,255 0.8 1,419 0.9
* 2000 is walk + bike, 2004 is walk only
The county estimates program which is used to weight the ACS data is
drawing considerable fire, as evidenced by the post by Jeffrey P. Levin
from the City of Oakland.
Things for State DOTs and MPOs to consider if you feel that the ACS
will not provide you with quality data:
1. Can your organization leverage enough political resources to bring
back a Census "long form" ?
2. Would improvements to the county estimates program make you feel
more comfortable with the ACS results?
3. Should your organization consider conducting a very large sample
survey, similar to the surveys conducted in the 1950's and 1960's where
sample sizes of 3 - 5 % of all households were asked to completed a
travel diary? One of the goals of these surveys was to produce an O/D
matrix for a limited number of zones. An area with 1 million population
might have 400,000 households, therefore a 4% sample would be 16,000
households. Let's estimate the cost of a household survey at a
conservative $150 per complete, resulting in a estimate of $2.4 million.
Keep in mind that the response rates to recent regional household
travel surveys have been between 25-30%, which is much lower than the
ACS, thus, risking much higher sample bias. Once you get the results,
you will need to determine a method to weight your results for regional
totals.
4. Should your organization implement a survey on group quarters
population, or do you believe that the ACS will include group quarters
in 2006, as planned.
5. Should you find an alternative data source for home-to-work flows.
Sorry for the long post, and hope that my table comes over without
distortion.
Elaine
FHWA Office of Planning
-----Original Message-----
From: ctpp-news-bounces(a)chrispy.net
[mailto:ctpp-news-bounces@chrispy.net]On Behalf Of Thabet Zakaria
Sent: Tuesday, September 13, 2005 6:54 AM
To: ctpp-news(a)chrispy.net
Subject: [CTPP] 2004 ACS Data Release
2004 ACS Data Release by Thabet Zakaria dated September 12, 2005 is
resubmitted in PDF.
_______________________________________________
ctpp-news mailing list
ctpp-news(a)chrispy.net
http://www.chrispy.net/mailman/listinfo/ctpp-news
- Richard Lin, Ph.D.
Demographer
Colorado Division of Local Government
(303)866-4989, fax (303)866-2660
richard.lin(a)state.co.us
_______________________________________________
ctpp-news mailing list
ctpp-news(a)chrispy.net
http://www.chrispy.net/mailman/listinfo/ctpp-news
After responding to the points raised by some of my colleagues last week
about the ACS 2004 data, I have received two more comments. The first one
is from Richard Lin, who is hoping that the quality of the ACS data by
Census 2010 will be equal or even better than that of the long-form. This
is not possible because the proposed larger sample size of 2005-2009 (Full
Nationwide Implementation) is about 70 percent of the sample size of the
long-form. Even if Congress approves funding for five years as requested by
the CB, the quality of the ACS data will be lower than the census long-form
because of many other reasons, including the ACS sample is weighted to an
estimated population rather than census, might not include group quarters
population, excludes the impacts of significant discontinuities in
population and employment characteristics, and is not conducted in one year.
In addition, the CB is promising to provide us with TAZ or Census Tract
level data after five year accumulation of ACS data. According to sampling
theory, the sampling errors in the ACS data at the zonal level will be much
greater than that at the county level. We all love to have accurate current
TAZ level data for our planning studies but this wont happen because of the
limitations of the proposed ACS methodology and data collection procedures.
Finally, I would like to ask this question. If you are working for the City
of Oakland, CA and you know that the City is growing, which population
number would you prefer for 2004: Census 2000 of 399,000 or ACS estimated
number of 365,000?
Glen Ahlert suggested to perform a liner regression on the annual ACS data
with respect to year for the areas that are undergoing continual population
or employment growth. Then interpolate or extrapolate to the year of
interest, than to average the annual ACS data. He also stated correctly
that regression analysis wont be reliable for areas that experienced
significant discontinuities in land use and transportation development. I
dont think the CB will accept such a method because it does not work for
the whole country.
For the majority of geographic areas, especially those expected to be
undergoing continual population growth or other demographic changes at a
relatively steady rate, it may be better to perform a linear regression
on the annual ACS data with respect to year, then interpolate or
extrapolate to the year of interest, than to average the annual ACS
data. Obviously there are other areas that have experienced significant
discontinuities in population or demographic characteristics for which
this method would not be very reliable, but then neither would using
multi-year ACS averages.
Glen Ahlert
Director
Lee County MPO
(239) 338-2550 ext. 224
Dear Mr. Lin- I disagree with you about census long form vs. ACS. I cannot use the ACS for any transportation planning analysis except maybe to see where general trends are going. Even there, looking at NJ data by county, there are some weird changes in mode data even from year to year. I use the 2000 Census data almost every week, and the ACS, unless it can get down to a smaller geography, does not do much good. Unless you are doing some high level comparisons about average travel time or average mode split etc., it cannot be used for what most transportation planners need, such as trip distribution, mode split, auto occupancy, vehicle ownership etc. If there is no long form, then we will have to take what we can from the ACS and combine it with other data, surveys etc.
-----Original Message-----
From: ctpp-news-bounces(a)chrispy.net [mailto:ctpp-news-bounces@chrispy.net] On Behalf Of Richard Lin
Sent: Friday, September 16, 2005 7:48 PM
To: ctpp-news(a)chrispy.net; tzakaria(a)dvrpc.org; Elaine.Murakami(a)fhwa.dot.gov
Subject: RE: [CTPP] 2004 ACS Data Release
Dear All,
We all realize the limitations and quality of the ACS data. Also, we all are seeking better resolutions. First of all, we hope the Census Bureau gets enough funding to carrying out ACS operations and to improving the quality of data by accumulating greater samples. Hope by Census 2010, the quality of the ACS data is equal to or even better than that of the traditional long form. Without annual ACS survey, we will fall back again to the once a 10-year long form data availability. Will you prefer still using the Census 2000 Long Form data in 2005 or do you appreciate the limited 2000-2004 ACS data? Without the annual ACS data, we will go back to the dark again.
The good news is that if the Census Bureau gets funding to continue ACS survey through Census 2010 then we should have much more reliable and larger sample ACS data for better quality time series.
With the sampling error variation, we should be more cautious about time series fluctuations and verify the true changes in numbers such as the Jew Jersey's Mercer County having lesser workers riding bikes or walk to work (5,450 or 3.5 in Census 2000 comparing to 2,924 or 1.9 in ACS 2004). We should also be cautious when we compare apple to orange. We may check the data at state (NJ) level by comparing the state's Census 2000 numbers and ACS 2000 numbers to establish a ratio factor for county adjustment for time series comparison.
>>> "Murakami, Elaine" <Elaine.Murakami(a)fhwa.dot.gov> 9/14/2005 1:39:12
PM >>>
Dear Zak,
Thank you for providing your opinion on the disutility of the ACS 2004 data for your applications, and I am glad that others are examining data for their area and are sharing their findings with the listserv.
My earlier email was sent to CAUTION people who might try to compare 2000 decennial data to ACS 2004 data. As I said in my earlier email, it is probably better to compare ACS results to ACS results, than to compare ACS results to decennial census results, however, results from ACS from 2000 are very limited. The geographic coverage of County level data from the ACS in 2000 is sparse, which makes comparisons for ACS between 2000 and 2004 possible only in some areas.
The ACS did not go into "full implementation" until 2005, and it will require multiple years of data collection before data for small areas will be reliable because of the small samples collected each month and
averaged over time. As Tom Marchwinski pointed out, averaging over
multiple years creates other problems. In 2005, the CB begin differential non-response follow-up, so that in areas (tracts) with low mail-back responses, there is a greater rate (1:2 and 2:5 instead of
1:3) made to follow-up non-respondents from the mail-back and CATI portion of the ACS. Not surprisingly, the areas with low mail-back responses are more likely to be low income, and higher shares of African American and Hispanic populations. My guess is that this will cause a shift in numbers between 2004 and 2005, and could impact variables such
as carpooling and transit use and number of vehicles in households.
The Census Bureau recommends that the ACS should be used to describe characteristics, and not to use it for COUNTS. When I examine the Mercer County NJ data comparing Census 2000 (workers in households) to ACS 2004 (workers in hhlds), I find that about the only thing to say is that "driving alone" appears to increase from 77 to 78%, and "carpooling" to decrease from 11 to 10%, "rail" also appears to increase from 4 to nearly 6%. (Worked at Home is not included in column in the table below). HOWEVER, this is not taking into account the effects of the different survey methods, where, generally speaking, the decennial census has a greater share reporting "carpooling," which is why a BRIDGE from decennial 2000 data is so important. Thus, it is probably incorrect to say that carpooling is declining.
Mercer County, NJ
Workers in Hhlds Census 2000 ACS 2004
Number Pct Number Pct
Total 153,665 153,041
Drove alone 118,390 77.0 119,597 78.1
2-person CP 13,105 8.5 12,026 7.9
3+ person CP 4,580 3.0 2,722 1.8
Bus/trolley bus 4,585 3.0 4,390
2.9
Streetcar/trolleycar+ 195 0.1 1,168 0.8
Railroad or ferry 6,105 4.0 8,795
5.7
Bike or Walk* 5,450 3.5 2,924 1.9
Taxi/motorcycle/other 1,255 0.8 1,419 0.9
* 2000 is walk + bike, 2004 is walk only
The county estimates program which is used to weight the ACS data is drawing considerable fire, as evidenced by the post by Jeffrey P. Levin from the City of Oakland.
Things for State DOTs and MPOs to consider if you feel that the ACS will not provide you with quality data:
1. Can your organization leverage enough political resources to bring back a Census "long form" ?
2. Would improvements to the county estimates program make you feel
more comfortable with the ACS results?
3. Should your organization consider conducting a very large sample survey, similar to the surveys conducted in the 1950's and 1960's where sample sizes of 3 - 5 % of all households were asked to completed a travel diary? One of the goals of these surveys was to produce an O/D matrix for a limited number of zones. An area with 1 million population might have 400,000 households, therefore a 4% sample would be 16,000 households. Let's estimate the cost of a household survey at a conservative $150 per complete, resulting in a estimate of $2.4 million. Keep in mind that the response rates to recent regional household travel surveys have been between 25-30%, which is much lower than the ACS, thus, risking much higher sample bias. Once you get the results, you will need to determine a method to weight your results for regional totals.
4. Should your organization implement a survey on group quarters population, or do you believe that the ACS will include group quarters in 2006, as planned.
5. Should you find an alternative data source for home-to-work flows.
Sorry for the long post, and hope that my table comes over without distortion.
Elaine
FHWA Office of Planning
-----Original Message-----
From: ctpp-news-bounces(a)chrispy.net
[mailto:ctpp-news-bounces@chrispy.net]On Behalf Of Thabet Zakaria
Sent: Tuesday, September 13, 2005 6:54 AM
To: ctpp-news(a)chrispy.net
Subject: [CTPP] 2004 ACS Data Release
2004 ACS Data Release by Thabet Zakaria dated September 12, 2005 is resubmitted in PDF.
_______________________________________________
ctpp-news mailing list
ctpp-news(a)chrispy.net
http://www.chrispy.net/mailman/listinfo/ctpp-news
- Richard Lin, Ph.D.
Demographer
Colorado Division of Local Government
(303)866-4989, fax (303)866-2660
richard.lin(a)state.co.us _______________________________________________
ctpp-news mailing list
ctpp-news(a)chrispy.net http://www.chrispy.net/mailman/listinfo/ctpp-news
Dear All,
We all realize the limitations and quality of the ACS data. Also,
we all are seeking better resolutions. First of all, we hope the Census
Bureau gets enough funding to carrying out ACS operations and to
improving the quality of data by accumulating greater samples. Hope by
Census 2010, the quality of the ACS data is equal to or even better than
that of the traditional long form. Without annual ACS survey, we will
fall back again to the once a 10-year long form data availability. Will
you prefer still using the Census 2000 Long Form data in 2005 or do you
appreciate the limited 2000-2004 ACS data? Without the annual ACS data,
we will go back to the dark again.
The good news is that if the Census Bureau gets funding to
continue ACS survey through Census 2010 then we should have much more
reliable and larger sample ACS data for better quality time series.
With the sampling error variation, we should be more cautious about time
series fluctuations and verify the true changes in numbers such as the
Jew Jersey's Mercer County having lesser workers riding bikes or walk to
work (5,450 or 3.5 in Census 2000 comparing to 2,924 or 1.9 in ACS
2004). We should also be cautious when we compare apple to orange. We
may check the data at state (NJ) level by comparing the state's Census
2000 numbers and ACS 2000 numbers to establish a ratio factor for county
adjustment for time series comparison.
>>> "Murakami, Elaine" <Elaine.Murakami(a)fhwa.dot.gov> 9/14/2005 1:39:12
PM >>>
Dear Zak,
Thank you for providing your opinion on the disutility of the ACS 2004
data for your applications, and I am glad that others are examining data
for their area and are sharing their findings with the listserv.
My earlier email was sent to CAUTION people who might try to compare
2000 decennial data to ACS 2004 data. As I said in my earlier email, it
is probably better to compare ACS results to ACS results, than to
compare ACS results to decennial census results, however, results from
ACS from 2000 are very limited. The geographic coverage of County level
data from the ACS in 2000 is sparse, which makes comparisons for ACS
between 2000 and 2004 possible only in some areas.
The ACS did not go into "full implementation" until 2005, and it will
require multiple years of data collection before data for small areas
will be reliable because of the small samples collected each month and
averaged over time. As Tom Marchwinski pointed out, averaging over
multiple years creates other problems. In 2005, the CB begin
differential non-response follow-up, so that in areas (tracts) with low
mail-back responses, there is a greater rate (1:2 and 2:5 instead of
1:3) made to follow-up non-respondents from the mail-back and CATI
portion of the ACS. Not surprisingly, the areas with low mail-back
responses are more likely to be low income, and higher shares of African
American and Hispanic populations. My guess is that this will cause a
shift in numbers between 2004 and 2005, and could impact variables such
as carpooling and transit use and number of vehicles in households.
The Census Bureau recommends that the ACS should be used to describe
characteristics, and not to use it for COUNTS. When I examine the
Mercer County NJ data comparing Census 2000 (workers in households) to
ACS 2004 (workers in hhlds), I find that about the only thing to say is
that "driving alone" appears to increase from 77 to 78%, and
"carpooling" to decrease from 11 to 10%, "rail" also appears to increase
from 4 to nearly 6%. (Worked at Home is not included in column in the
table below). HOWEVER, this is not taking into account the effects of
the different survey methods, where, generally speaking, the decennial
census has a greater share reporting "carpooling," which is why a
BRIDGE from decennial 2000 data is so important. Thus, it is probably
incorrect to say that carpooling is declining.
Mercer County, NJ
Workers in Hhlds Census 2000 ACS 2004
Number Pct Number Pct
Total 153,665 153,041
Drove alone 118,390 77.0 119,597 78.1
2-person CP 13,105 8.5 12,026 7.9
3+ person CP 4,580 3.0 2,722 1.8
Bus/trolley bus 4,585 3.0 4,390
2.9
Streetcar/trolleycar+ 195 0.1 1,168 0.8
Railroad or ferry 6,105 4.0 8,795
5.7
Bike or Walk* 5,450 3.5 2,924 1.9
Taxi/motorcycle/other 1,255 0.8 1,419 0.9
* 2000 is walk + bike, 2004 is walk only
The county estimates program which is used to weight the ACS data is
drawing considerable fire, as evidenced by the post by Jeffrey P. Levin
from the City of Oakland.
Things for State DOTs and MPOs to consider if you feel that the ACS
will not provide you with quality data:
1. Can your organization leverage enough political resources to bring
back a Census "long form" ?
2. Would improvements to the county estimates program make you feel
more comfortable with the ACS results?
3. Should your organization consider conducting a very large sample
survey, similar to the surveys conducted in the 1950's and 1960's where
sample sizes of 3 - 5 % of all households were asked to completed a
travel diary? One of the goals of these surveys was to produce an O/D
matrix for a limited number of zones. An area with 1 million population
might have 400,000 households, therefore a 4% sample would be 16,000
households. Let's estimate the cost of a household survey at a
conservative $150 per complete, resulting in a estimate of $2.4 million.
Keep in mind that the response rates to recent regional household
travel surveys have been between 25-30%, which is much lower than the
ACS, thus, risking much higher sample bias. Once you get the results,
you will need to determine a method to weight your results for regional
totals.
4. Should your organization implement a survey on group quarters
population, or do you believe that the ACS will include group quarters
in 2006, as planned.
5. Should you find an alternative data source for home-to-work flows.
Sorry for the long post, and hope that my table comes over without
distortion.
Elaine
FHWA Office of Planning
-----Original Message-----
From: ctpp-news-bounces(a)chrispy.net
[mailto:ctpp-news-bounces@chrispy.net]On Behalf Of Thabet Zakaria
Sent: Tuesday, September 13, 2005 6:54 AM
To: ctpp-news(a)chrispy.net
Subject: [CTPP] 2004 ACS Data Release
2004 ACS Data Release by Thabet Zakaria dated September 12, 2005 is
resubmitted in PDF.
_______________________________________________
ctpp-news mailing list
ctpp-news(a)chrispy.net
http://www.chrispy.net/mailman/listinfo/ctpp-news
- Richard Lin, Ph.D.
Demographer
Colorado Division of Local Government
(303)866-4989, fax (303)866-2660
richard.lin(a)state.co.us
I want to thank everyone who commented on my memo about the 2004 ACS Data
Release dated September 9, 2005. Also, I would like to respond and clarify
some of the points raised by my colleagues throughout the country. As you
know, the issues I raised in my September 9th post are not new; they were
raised two years ago in Washington, DC at the TRB Annual Conference and at
the Census Data for Transportation Conference in Irvine, CA, May 2005.
Ed Christopher asked for help from the list server and census folks who
produced the 2004 population estimates and 2004 ACS estimates. We got help
from some of the list servers. But I dont think we are going to have any
response from the census folk because their evaluation of the ACS data
indicated conclusions similar to mine (see for example Census Report No. 5
on the results of ACS). They may argue that the quality of the ACS product
depends on the methodology, assumptions, and resources available. I think
they are doing their best to obtain quality data from a small and
unrepresentative sample. The Guide to Using the ACS and the NCHRP
Project are probably not going to change the quality of ACS data much,
which results from the current ACS assumptions, sample size, and content.
Elaine Murakami - I completely agree with you that it is better to compare
ACS results to ACS results rather than Census 2000. I also agree with you
that the 2005 data will be more accurate than ACS 2004. The problem is that
neither ACS 2004 nor ACS 2005 data are acceptable for transportation
planning, especially for variables with small values such as walk and
transit modes. The margin of error in these mode estimates is very large
due to sample size and other factors. Your table for Mercer County is
correct. But I think it is better to compare Census 2000 with ACS 2000
(apples to apples). Also, your conclusions on mode characteristics seems to
be reasonable. As you know, we need TAZ level data for our studies. The
current ACS program will produce larger errors than those estimated at the
county level.
Of all the things that you outlined for alternatives to replace the ACS, I
would say bring back the long-form. I know it is not easy to reverse the
decision that was made by the CB two years ago. Although not perfect, the
responses to the long-form will be more reliable because it will be
conducted with the decennial census, which is respected by almost everyone
in the country. I think that FHWA should take the lead in promoting this
effort and tell the country that the ACS data are not acceptable for the
transportation community since the margin of error is very large. I really
dont know the effective channels of communication to have the long-form
back in Census 2010. But I think the ACS issues should be brought to the
attention of decision makers within AASHTO, APTA, NARC, TRB, Public Data
Users, 2010 Census Advisory Committee, AMPO, 2010 Census Advisory Committee,
AMPO, House Subcommittee on the Census, and other organizations. Of course
political pressure is always needed. The other options you outlined are not
really alternatives to the long-form. Improvements to the ACS county
program will help a little but wont solve the problem. Large surveys
similar to the surveys conducted in the 1950's and 1960's are out of the
question.
Only the CB could implement a survey on group quarters population.
Including group quarters population will improve the ACS quality but wont
solve the major problems in ACS data due to sample size. Home-to-Work flows
is not the only data item we obtain from the long-form.
I dont think the CB will be able to improve its county population estimates
significantly within the available resources. More than 10 years ago, I
raised this question of quality with the CB and was told We are only
preparing county estimates, not a census. Yet they are now using the
county estimates instead of the census counts.
I completely agree with Patty Becker that including group quarters
population wont solve the whole problem and the average of 2003-2007 in
New Orleans will not be good. Also, I agree with Patty that it is good to
have one year data for places of 65,000 or more population and 3-year data
sets for places of 20,000 or more. Unfortunately, the current ACS program
is not expected to produce accurate data for such areas because of sample
size and many other factors.
Tom Marchwinski gave excellent reasons for the error in ACS products. He
questions the validity of aggregating five years of data and calling it a
five-year average. He indicated that opening a major transit service, gas
prices, major land use development, etc., can skew the ACS results and
introduce additional errors in the ACD data.
Grey Lipton stated incorrectly that Mercer County population declined by
11,000 during between the two surveys (about 9,000 if you use the 2000
census population). According to census population estimates, Mercer
County population grew by 13,693 persons between July 2000 and July 2004.
Also, his conclusion about the number of walkers and the sample error is
erroneous because it is based on the wrong numbers.
In summery, the errors in the county ACS data for 2000-2004 are very large
and the data cannot be used for transportation planning. Like 2004, the
error in the 2005 ACS (Full Nationwide Implementation) for areas 65,000 plus
population is expected to be large because of many reasons, including the
ACS sample is smaller than decennial census, does not include group quarters
population, and is weighted to an estimated population rather than census
counts. The CB cannot do much to improve the quality of ACS because of
limited resources. In order to obtain quality data for transportation
planning, we have no option but to bring back the census long-form and
include it in Census 2010. Although not perfect, decennial census data are
the best estimates for all the variables that we need for transportation
studies, including population, households, car ownership, travel modes
including carpools, employed person by place of residence and place of work,
Journey-to-Work, income, and travel time by mode of travel. FHWA is up to
the challenge because it has been a leader in this area for the past 50
years.
SENATE APPROVES COMMERCE FUNDING BILL;
NO FUNDS ADDED FOR CENSUS BUREAU
After a week of deliberations, the U.S. Senate approved the
appropriations bill that will pay for Census Bureau programs in fiscal
year 2006, clearing the way for a conference with the House of
Representatives in the coming weeks to iron out substantial differences
between the two versions of the bill. There were no changes to the
Census Bureaus funding level set by the Senate Appropriations Committee
in June. The new fiscal year begins October 1, 2005.
The Senate passed the massive FY06 Commerce, Justice, and Science
Appropriations bill (H.R. 2862) by a vote of 91 - 4 on September 15. In
her opening statement on the floor, Sen. Barbara Mikulski (D-MD),
ranking minority member on the Commerce spending subcommittee, called
the Census Bureau one of the flashing yellow lights in the bill,
referring to the funding level of $727.4 million, which is $150 million
below the Presidents request and $17 million less than 2005. She
highlighted the importance of the agencys work to continually
evaluat[e] who we are and where we live and give[s] us important
information so that communities and businesses can develop everything
from transportation and education plans to business targeting new
demography in our country. The committee believed the Census Bureau
could operate next year with a modest cut from this years budget,
Sen. Mikulski acknowledged. We hope to restore that in conference,
she said, although it is not clear whether the senator was referring to
the 2005 funding level or the $85 million-higher figure the House
approved. Sen. Mikulski also cautioned, however, that in the aftermath
of Hurricane Katrina, [w]e will be reevaluating as we go along. The
Commerce appropriations bill also covers the National Oceanic and
Atmospheric Administration (NOAA), Economic Development Administration,
and Small Business Administrations disaster loans program. Senators
approved a number of amendments to assist hurricane victims and to
increase funding for law enforcement and anti-drug programs.
The Office of Management and Budget issued its Statement of
Administration Policy (SAP) last week as the Senate prepared to consider
the Commerce-Justice-State spending bill. While noting that funding for
the Commerce Department was $1.1 billion above the Presidents request,
the Administration expressed concern about funding for the Census
Bureau. The SAP confirmed that, under the Senate funding level, the
bureau would suspend the American Community Survey, increase the
lifecycle cost of the 2010 census, and lead to a less accurate
Census. The letter to the Senate noted that timely and accurate
Census data are necessary in recovery efforts for crisis situations
(emphasis added). The Senate funding level also would jeopardize the
accuracy of the national income accounts and lead to the elimination of
important economic data series, OMB said. The full statement from OMB
is available at
http://www.whitehouse.gov/omb/legislative/sap/109-1/hr2862sap-s.pdf.
All members of the Senate Commerce-Justice-Science Appropriations
subcommittee, as well as the chairman and senior Democrat of the full
committee, were appointed as conferees on the bill. It is too early to
know how quickly House and Senate negotiators will meet.
Stakeholders advocate for census programs: A diverse group of census
stakeholders, representing business and industry, civil rights and child
advocates, data users, and others, sent a second letter to the Senate
last week, urging lawmakers to increase funding for the American
Community Survey (ACS), 2010 census planning, and other important
statistical programs. Organized by the Communications Consortium Media
Center as The Census Project, the stakeholders noted that the Senate
funding level would jeopardize continuation of the ACS and the 2006
Census Test in Travis County, TX, and on the Cheyenne River Sioux Indian
Reservation. A copy of this and previous letters are posted at
www.thecensusproject.org. (Note: Several organizations asked to sign
the letter after it was sent; their names appear on the permanent copy
posted on the web site.)
Other stakeholders, including the U.S. Chamber of Commerce, JCPenney,
and the chairman of the Cheyenne River Sioux tribe, have sent their own
letters to senators, encouraging full funding for the ACS, 2010 census,
and economic statistical programs endangered by potential funding cuts.
New census advisory committee named: The Census Bureau has selected
members of the new 2010 Census Advisory Committee, the successor to the
Decennial Census Advisory Committee, which was disbanded last winter.
The committee will hold its first meeting on October 27-28, following an
orientation for organizations that were not members of the previous
committee.
The following organizations accepted invitations to join the advisory
panel. An asterisk indicates groups that did not serve previously.
American-Arab Anti-Discrimination Committee
American Farm Bureau Federation*
American Foundation for the Blind
Association of MultiEthnic Americans
Association of Public Data Users
California Rural Legal Assistance, Inc.
Consortium of Social Science Associations*
Council of Professional Associations on Federal Statistics
Election Data Services, Inc.*
Federal Reserve Bank of New York*
Housing Statistics Users Group
National Asian Pacific American Legal Consortium
National Association of Counties
National Association of Latino Elected and Appointed Officials
Educational Fund
National Association of Towns and Townships
National Conference of Black Mayors*
National Congress of American Indians
POLIDATA Political Data Analysis*
U.S. Chamber of Commerce
Change in prisoner count picks up support: The U.S. Conference of Mayors
adopted a resolution urging Congress to change the way prisoners are
counted in the decennial census. The resolution, approved at the
groups annual meeting in Chicago in June, calls for legislation to
require that prison inmates be counted at their pre-incarceration home.
Prisoners are counted at their place of incarceration under current
census residence rules.
The current method, the Mayors contend, draws a disproportionate amount
of funding to prison community locations, while restricting the amount
of funds that go to the actual home communities of the inmates. They
noted that inmates families often require services in the home
communities and that 95 percent of prisoners will return to those
communities upon release.
In June, the House Appropriations Committee directed the Census Bureau
to evaluate a change in the way prisoners are counted in the decennial
census and to issue a report within 90 days of enactment of the
appropriations bill. Senate appropriators did not include a similar
requirement in their report accompanying the Census Bureaus fiscal year
2006 funding bill.
Census News Briefs are prepared by Terri Ann Lowenthal, an independent
consultant in Washington, DC, with support from The Annie E. Casey
Foundation and other organizations. Ms. Lowenthal is also a consultant
to The Census Project, sponsored by the Communications Consortium Media
Center. All views expressed in the News Briefs are solely those of the
author. Please direct questions about the information in this News
Brief to Ms. Lowenthal at 202/484-3067 or by e-mail at
TerriAnn2K(a)aol.com. Please feel free to circulate this document to
other interested individuals and organizations.
--
Ed Christopher
708-283-3534 (V)
708-574-8131 (cell)
FHWA RC-TST-PLN
19900 Governors Dr
Olympia Fields, IL 60461
Dear Zak,
Thank you for providing your opinion on the disutility of the ACS 2004 data for your applications, and I am glad that others are examining data for their area and are sharing their findings with the listserv.
My earlier email was sent to CAUTION people who might try to compare 2000 decennial data to ACS 2004 data. As I said in my earlier email, it is probably better to compare ACS results to ACS results, than to compare ACS results to decennial census results, however, results from ACS from 2000 are very limited. The geographic coverage of County level data from the ACS in 2000 is sparse, which makes comparisons for ACS between 2000 and 2004 possible only in some areas.
The ACS did not go into "full implementation" until 2005, and it will require multiple years of data collection before data for small areas will be reliable because of the small samples collected each month and averaged over time. As Tom Marchwinski pointed out, averaging over multiple years creates other problems. In 2005, the CB begin differential non-response follow-up, so that in areas (tracts) with low mail-back responses, there is a greater rate (1:2 and 2:5 instead of 1:3) made to follow-up non-respondents from the mail-back and CATI portion of the ACS. Not surprisingly, the areas with low mail-back responses are more likely to be low income, and higher shares of African American and Hispanic populations. My guess is that this will cause a shift in numbers between 2004 and 2005, and could impact variables such as carpooling and transit use and number of vehicles in households.
The Census Bureau recommends that the ACS should be used to describe characteristics, and not to use it for COUNTS. When I examine the Mercer County NJ data comparing Census 2000 (workers in households) to ACS 2004 (workers in hhlds), I find that about the only thing to say is that "driving alone" appears to increase from 77 to 78%, and "carpooling" to decrease from 11 to 10%, "rail" also appears to increase from 4 to nearly 6%. (Worked at Home is not included in column in the table below). HOWEVER, this is not taking into account the effects of the different survey methods, where, generally speaking, the decennial census has a greater share reporting "carpooling," which is why a BRIDGE from decennial 2000 data is so important. Thus, it is probably incorrect to say that carpooling is declining.
Mercer County, NJ
Workers in Hhlds Census 2000 ACS 2004
Number Pct Number Pct
Total 153,665 153,041
Drove alone 118,390 77.0 119,597 78.1
2-person CP 13,105 8.5 12,026 7.9
3+ person CP 4,580 3.0 2,722 1.8
Bus/trolley bus 4,585 3.0 4,390 2.9
Streetcar/trolleycar+ 195 0.1 1,168 0.8
Railroad or ferry 6,105 4.0 8,795 5.7
Bike or Walk* 5,450 3.5 2,924 1.9
Taxi/motorcycle/other 1,255 0.8 1,419 0.9
* 2000 is walk + bike, 2004 is walk only
The county estimates program which is used to weight the ACS data is drawing considerable fire, as evidenced by the post by Jeffrey P. Levin from the City of Oakland.
Things for State DOTs and MPOs to consider if you feel that the ACS will not provide you with quality data:
1. Can your organization leverage enough political resources to bring back a Census "long form" ?
2. Would improvements to the county estimates program make you feel more comfortable with the ACS results?
3. Should your organization consider conducting a very large sample survey, similar to the surveys conducted in the 1950's and 1960's where sample sizes of 3 - 5 % of all households were asked to completed a travel diary? One of the goals of these surveys was to produce an O/D matrix for a limited number of zones. An area with 1 million population might have 400,000 households, therefore a 4% sample would be 16,000 households. Let's estimate the cost of a household survey at a conservative $150 per complete, resulting in a estimate of $2.4 million. Keep in mind that the response rates to recent regional household travel surveys have been between 25-30%, which is much lower than the ACS, thus, risking much higher sample bias. Once you get the results, you will need to determine a method to weight your results for regional totals.
4. Should your organization implement a survey on group quarters population, or do you believe that the ACS will include group quarters in 2006, as planned.
5. Should you find an alternative data source for home-to-work flows.
Sorry for the long post, and hope that my table comes over without distortion.
Elaine
FHWA Office of Planning
-----Original Message-----
From: ctpp-news-bounces(a)chrispy.net
[mailto:ctpp-news-bounces@chrispy.net]On Behalf Of Thabet Zakaria
Sent: Tuesday, September 13, 2005 6:54 AM
To: ctpp-news(a)chrispy.net
Subject: [CTPP] 2004 ACS Data Release
2004 ACS Data Release by Thabet Zakaria dated September 12, 2005 is
resubmitted in PDF.
I think there is a major problem in the ACS survey data for 2004 because of not including group quarters. Princeton University is in Mercer County, and a good part of its student population is housed in dorms or eating groups (fraternities), which are group quarters. Something like 40% of the work trips in Princeton Boro are walk to work, probably mostly univeristy related jobs, and students employed within the Boro. The ACS leaving out group quarters may explain part of this. I also have a problem with aggregating 5 years of data and calling it a 5 year average. NJT opened a 33 mile Light Rail line in early 2004, connecting Trenton to Camden. I noticed in the Burlington County data a big jump in "railroad" work trips, probably related to this new rail line. How does a major change in transit service get averaged with 3 other years when it is a permanent change. This applies to any region where a major change in transit service gets implemented. What about possible shifts in carpooling this year due to gas prices. A 5 year average may not be worth much. I think Dr. Zakaria raises some real issues here, especially with big error rates, and I think major changes in transit service, gas prices etc. can skew things.
-----Original Message-----
From: ctpp-news-bounces(a)chrispy.net [mailto:ctpp-news-bounces@chrispy.net] On Behalf Of Thabet Zakaria
Sent: Tuesday, September 13, 2005 9:54 AM
To: ctpp-news(a)chrispy.net
Subject: [CTPP] 2004 ACS Data Release
2004 ACS Data Release by Thabet Zakaria dated September 12, 2005 is resubmitted in PDF.
The population in the Mercer County declined by 11,000 during between the 2
surveys (about 9,000 if you use the 2000 Census population). This could
explain some of the difference. However, since the confidence intervals
overlap for the number of walkers, there is not a statistically significant
difference at the alpha =0.10 level. Sample error appears to be the
culprit. If alpha =0.05 was used, there would likely be a statistical
difference.
Greg Lipton
King County Transit
Seattle, WA
206 263-4673
-----Original Message-----
From: ed christopher [mailto:edc@berwyned.com]
Sent: Tuesday, September 13, 2005 1:02 PM
Cc: ctpp-news(a)chrispy.net
Subject: Re: [CTPP] 2004 ACS Data Release
Zak raises a lot of issues and I am hopefully that some of the Census folks
and
others on this list may be able to help. Needless to say, there is a whole
lot of
education that needs to be done when working with these data. Especially
when
dealing with year to year changes.
One area that particularly struck me in Zaks work was the comparison of ACS
2000
Mercer county walkers to ACS 2004 Mercer county walkers. When I saw Zak's
numbers I
went to Fact Finder to see for myself what may have been happening and this
is what
I found. First, Mercer county was not an ACS test site in 2000 so I used the
C2SS
data for 2000. Looking at those numbers I got 5,567 walkers (with a low of
2,465
and a high of 8,669) for 2000 and 2,924 (with a low of 1,556 to 4,292) for
2004. Or
put another way 2000 was 5,567 plus or minus 56% while 2004 was 2,924 plus
or minus
47%. Becoming even more curious, I checked out 2003 and found 4,303 walkers
(with a
low of 2,223 to 6,383). That turned out to be 4,303 walkers plus or minus
48%. So
what does all this mean? I am not quite sure except to say that for 2003
the
overall mode split of walkers was 2.6% with a range of 1.4% to 3.7%. The
mode split
of walkers for 2004 was 1.8% with a range of 1.0% to 2.6%. However you
slice it,
something is going on but just what???.
Hopefully, the "Guide to Using the ACS" that the Census Bureau is developing
and the
NCHRP project to develop a guidebook on using the ACS data will help us plow
though
the forest that is apparently growing right before our eyes.
Thabet Zakaria wrote:
> 2004 ACS Data Release by Thabet Zakaria dated September 12, 2005 is
> resubmitted in PDF.
>
> ------------------------------------------------------------------------
> Name: 2004 ACS Data Release.pdf
> 2004 ACS Data Release.pdf Type: Portable Document Format
(application/pdf)
> Encoding: base64
>
> ------------------------------------------------------------------------
> _______________________________________________
> ctpp-news mailing list
> ctpp-news(a)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(a)chrispy.net
http://www.chrispy.net/mailman/listinfo/ctpp-news
