Presentation is loading. Please wait.

Presentation is loading. Please wait.

Data issues with the American Community Survey

Similar presentations


Presentation on theme: "Data issues with the American Community Survey"— Presentation transcript:

1 Data issues with the American Community Survey
Elaine Murakami, FHWA CTPP for 1990 and 2000 was implemented as an AASHTO pooled fund, with a single contract between AASHTO and the CB. It has been very much a cooperative effort with USDOT, CB and AASHTO. As the CTPP2000 production winds down, we are transitioning our attention to ACS. A couple of things have happened since my last public presentation on ACS, at the TRB Applications conference in Portland, OR. First, TRB Census conference was held in May. And second, AASHTO SCOP has established a Census data working group. AMPO Travel Modeling Work Group, September 8, 2005

2 First, the good news… CB has added several transportation-related tables to their planned set of ACS tabulations. See handouts. This includes Residence tabulations of households such as: household size * # of vehicles A FEW tables by Place of Work The MORE that transportation and JTW questions are incorporated into the standard set of tabulations, the FASTER the data is available. (We don’t have to wait for special tab). Spreadsheet is available on the TRB Census Subcommittee webpage: HANDOUTS: Population Profile for Workers at Place of Residence Subject Table (Commuting) by Place of Residence Subject Table (Means of Transportation) by Place of Residence JTW Base Tables by Place of Residence (includes universe: “ households”, not just “workers”) Subject Table (Means of Transportation) by Place of Work

3 More positive things about ACS
The ACS will provide reliable data for large geographic units on an annual basis. The flip side is that the data for small areas will be compromised, with much larger standard errors, and the risk of the inability to tabulate small area (TAZ) home-to-work flows. One of the handouts is the multiyear profile for the U.S. total. Keep in mind that the ACS sample does not include group quarters. To me, one of the most surprising results of the 2000 to 2004 trend is the on-going decline in carpooling for work, and increasing shares of driving alone. We will have to see if this trend reverses with the recent increases in gasoline prices. Another thing to keep in mind is that the ACS was not in full implementation until 2005, and previous to that, only 1205 counties (out of over 3500) were in the sample. These 1205 counties are biased toward large metropolitan areas. There is likely to be a jump between 2004 and 2005 results.

4 Questions for today’s discussion
Do you want home-to-work flow (Part 3) from ACS if TAZs are larger, or TAZ-groups need to be used? Should AASHTO to sponsor a pooled fund for a CTPP from ACS? AASHTO SCOP has established Census Working Group, chaired by Jonette Kreideweis at MN DOT.

5 Key issues for Transportation
Sample size / geographic level of reporting / data quality / error Thresholds for FLOW tabulation CB Disclosure Review Board

6 Sample Size and Unweighted Responses
The number of addresses in the ACS sample is about 12.5 percent (over 60 months), compared to 16.6 percent in the decennial Long Form. The number of unweighted records in the ACS sample is 40 to 60 percent that of the decennial Long Form. Non-response follow-up surveys to ACS is a range of 1:2, 2:5 and 1:3 samples.

7 Implications of smaller number of completed surveys
Kiss TAZ-to-TAZ data GOODBYE, or define larger TAZs for Census tabulation. Need a geographic unit maybe a “Tract Group” or a TAZ-Group that is as large as 2 or 3 tracts together, especially for home-to-work flow data. TAZs may be different from TAZs defined in 2000, and may be more like the size of a census tract, or even a combination of census tract (for flow reporting).

8 Thresholds required for CTPP2000 for flow tabulation
Thresholds are based on unweighted records. The smaller number of completed records in ACS results in a loss of O/D pairs meeting the threshold.

9 Impact of Thresholds Comparing Decennial LF to ACS Part 3 Broward County, FL Tract-to-Tract
Pt 3 WITHOUT Thresholds Pt 3 WITH Thresholds Part 1 total workers # of Pairs Total Workers LF 36,300 777,900 8,200 424,700 742,600 ACS 21,500 616,500 2,800 237,200 747,400 FROM TRACT-to TRACT level data: For Broward County, the 3 record threshold using the 2000 decennial census long form data results in 55 percent of the workers being counted (compared to no threshold). When we use the ACS test data, we find that only 30 percent of the total worker flows compared to 2000 long form without a threshold. Some of difference is due to smaller # of records and subject to thresholds, but some of the difference is due to ACS not undergoing “extended allocation.” This largely affects the number of workers accounted for in the flow pairs, rather than the total # of O/D pairs. When the thresholds are applied, the number of flow pairs at the tract-to-tract level is reduced from 8,200 pairs to 2,800 pairs using ACS. 22% of O/D pairs 30% of workers 8% of O/D Pairs

10 Does 60 months of accumulation result in reduced risk of disclosure
Does 60 months of accumulation result in reduced risk of disclosure? We think “yes” People move residence People move workplace location People change means of transportation to work. People change departure time. People change # of vehicles in hhld. Current ACS– 25 percent of workplace locations are imputed/allocated. Need to examine PANEL data to determine RATES! For the people in this room, raise your hand if you moved your residence in the last 5 years. Keep your hands up. Now, if you moved your workplace location, raise your hand.

11 What rules will the DRB impose?
Will the DRB impose thresholds for flow tabulation? We hope not! If the threshold is 3 for key tables, as in 2000, will transportation planners find it useful or not? Meeting scheduled with Laura Zayatz, chair of the CB’s DRB on Sept 20. Find out if “income” is the problem and what might be done as an alternative. CB website say “thresholds” will be applied.

12 What do we want from the CB?
Alternatives to rounding and thresholds for disclosure avoidance. Improved allocation procedures. Want to continue working with CB on allocation and imputation of means of transportation, # of vehicles, and place of work geocoding. Believe that quality of data from ACS is high, especially with 1:3 NRFU surveys. We would like to see more research on the potential of synthetic data as a disclosure avoidance technique.

13 How should we augment ACS data?
Explore alternatives for Home-to-Work Flow matrices, including LEHD Consider workplace surveys NPTS for 2007, for trip length distribution curves for various trip purposes/activities New technology approaches for O/D matrices (GPS, RFID, etc) Right now, the next NPTS is not yet funded. As of today, the goal is to implement in 2007, and use the same questionnaire and methods as 2001 survey.

14 Questions for today’s discussion
Do you want home-to-work flow (Part 3) from ACS ? Should AASHTO sponsor a pooled fund for a CTPP from ACS? Maybe there are too many problems to consider a small area flow tabulation, and we should just take the CB standard table set and not do an AASHTO pool. Next few slides are Elaine’s thoughts on issues to be addressed by AASHTO SCOP Census Working Group.

15 AASHTO pooled fund Potential concepts: Geography
New TAZ definition in 2008, input to TIGER New geographic unit: “TAZ-Groups” for Flow tabulation?

16 AASHTO pooled fund Potential concepts: Table content
Special Tables for large geographic units starting in 2007? 2008? Or, are the standard tabulations by CB sufficient? Tables for areas with complex transit, where “bus” and “other transit” is insufficient. Tables with Income using TAZ-Group, and eliminating Income as a variable for TAZ flows.

17 AASHTO pooled fund Potential concepts: Add a research component
Alternatives to thresholds for disclosure avoidance. Alternatives to “income” as a tabulation variable. Improving business locations (esp. multisite businesses) used in the LEHD program. Methods to combine ACS with LEHD (administrative records).

18 Last remarks TRB Conference on Census and ACS should have final report available on web by end of the year. NCHRP ACS Guidebook should be complete by end of the year Please complete the AASHTO CTPP User Survey at MPO representatives on the TRB Committee include: Ken Cervenka (NCTCOG), Dunbar Brooks (Baltimore Met Council), Tom Kane (Des Moines MPO), Chuck Purvis (MTC), Jeff Tayman (SANDAG). NCHRP contractor is Cambridge Systematics, with Kevin Tierney as the PI


Download ppt "Data issues with the American Community Survey"

Similar presentations


Ads by Google