Use of the U.S. Census Bureau’s Public Use Microdata Sample (PUMS) by State Departments of Transportation and Metropolitan Planning Organizations (2012) / Chapter Skim
Currently Skimming:
Pages 28-54
The Chapter Skim interface presents what we've algorithmically identified as the most significant single chunk of text within every page in the chapter.
Select key terms on the right to highlight them within pages of the chapter.
From page 28... ...
28 CHAPTER FOUR APPLICATIONS OF THE PUBLIC USE MICRODATA SAMPLE DATA COMMON TRANSPORTATION PLANNING USES OF CENSUS DATA Transportation planners use Census data to support a wide range of functions and planning activities. Understanding the relationships between household and population characteristics information collected by the Census Bureau and transportation system usage data is a key aspect of transportation planning.
|
From page 29... ...
29 up-to-date CTPP-like data tables, albeit with less precision in the estimates and less geographic detail. • Disaggregate analyses – Planners and modelers frequently require household- or person-level (disaggregate)
|
From page 30... ...
30 FHWA (2002) provided two specific examples of how PUMS data have helped in environmental justice analyses: • For a study in Atlanta, PUMS data helped labor market and transportation analysts distinguish entry-level jobs and other occupations most suitable to persons with limited formal education or training.
|
From page 31... ...
31 FIGURE 14 An illustration of PUMS cross-tabulation analyses. Source: Based on Purvis (2011)
|
From page 32... ...
32 Censuses. However, the available Census tabulations at the tract level did not allow the authors to analyze people by age, job type, and income, so they relied on the year 1990 and year 2000 5% PUMS data for 92 PUMAs for these multiway classifications.
|
From page 33... ...
33 to be the result of lower housing vacancy rates in the growth areas, rather than the result of redevelopment that favors transit-friendly subareas.
|
From page 34... ...
34 To weight the survey results to account for deliberate and circumstantial oversampling and undersampling, analysts factor all the survey results with a specific characteristic by the ratio of the sum of the actual households with the characteristic (taken from Census-based estimates) to the sum of survey records with the characteristic.
|
From page 35... ...
35 – PUMS employment status variable categorized into one dummy variable (Employed) , – Adult student and adult full-time student dummy variables created based on age, student, and employed dummy variables, and – All person-based dummy variables summed by household, and sums merged to household records.
|
From page 36... ...
36 One issue that has arisen with using PUMS data to expand travel surveys is that estimates derived from the PUMS data may be somewhat different than estimates based on other Census data products. Therefore, the application of weights to have travel surveys better match PUMS data summaries may lead to survey results that are inconsistent with other summaries of Census data.
|
From page 37... ...
37 MWCOG used PUMS commuter data to perform validation tests on its household survey data (R. Griffiths, personal communication, Apr.
|
From page 38... ...
38 MPO planning area boundaries. It is much simpler, in their view, to work with available data tabulations for smaller Census geographic delineations that better correspond to the MPO geographies (MPO area and traffic analysis zone structure)
|
From page 39... ...
39 data from a comprehensive household travel survey. Baber (2004)
|
From page 40... ...
40 history of advanced travel demand models, beginning with trip-based models and proceeding to tour-based models, TRANSIMS (TRansportation ANalysis SIMulation System) , and activity-based travel demand models (Vovsha et al.
|
From page 41... ...
41 including household size, age, gender, household income, level of educational attainment, and race/ethnicity (Morris and Smart 2011)
|
From page 42... ...
42 sity in the workplace subregion. The multiyear PUMS data set with the merged spatial information included more than 150,000 workers, about 35,000 of whom were foreign-born (Chatman and Klein 2011)
|
From page 43... ...
43 • Represent an entire day of activities and travel for each member of a synthetic population, using stochastic microsimulation; • Consist of an integrated system of econometric models; and • Include traditional traffic and public transport assignment components. With regard to the PUMS data usage, the key point is that all the activity-based model systems rely on microsimulation-based model application, and therefore also rely on synthetic population modules.
|
From page 44... ...
44 FIGURE 17 General structure of regional activity-based travel demand model systems. Source: Mark Bradley Research and Consulting and J.L.
|
From page 45... ...
45 As household travel survey data collection has become more difficult and expensive, transportation researchers have become more interested in developing simulated or model-based survey data. A goal of this research has been to develop ways to combine local socio-demographic data for individuals/households (from sources such as Census Bureau data)
|
From page 46... ...
46 without the application of microsimulation techniques. Tripbased models and many types of traditional land use models are applied through the calculation of fractional probabilities to aggregate segments of households at each step of a model system.
|
From page 47... ...
47 ing discussion focuses on that approach. A brief discussion of the CO approach is included at the end of this section.
|
From page 48... ...
48 other Census sources, synthesizer software needs to include a significant amount of computer code to address potential issues. One planner suggested that the Census Bureau consider the disclosure ramifications of providing PUMS data for a greater percentage of ACS records, or even all ACS records (J.
|
From page 49... ...
49 Axhausen (2011)
|
From page 50... ...
50 are using the PUMS data to establish the base-year synthetic population and to set the probabilities of certain events such as births, migration, marriage, and divorce. The synthesized population is then modified on a year-by-year basis to develop synthesized population forecasts.
|
From page 51... ...
51 Müller and Axhausen (2011) raise some concern about the development of alternative synthesizers, rather than a single best one: Given the difficulties that routinely arise when trying to properly create a synthetic population, it seems worthwhile to invest time to develop a generic software solution.
|
From page 52... ...
52 zone. Statistics are calculated to measure the fit of the subset to the known marginal distributions of control variables in the zone.
|
From page 53... ...
53 TABLE 21 SUMMARY OF PUMS DATA USES SUMMARIZED IN THIS SYNTHESIS PUMS Data Usage Reasons for Using PUMS Data/Benefits of PUMS Noted by Planners PUMS Drawbacks Noted by Planners Houston-Galveston Area Transportation Profile (Ju 2007) Comparison of PUMS variables over time.
|
From page 54... ...
54 • Year-to-year inconsistencies in PUMS files – The ACS has changed some questions since its introduction. These changes are reflected in the PUMS data files, so data users must match data dictionaries to analyze successive years.
|
Key Terms
This material may be derived from roughly machine-read images, and so is provided only to facilitate research.
More
information on Chapter Skim is available.