1 / 45

Keeping Up with the New Census Data

Keeping Up with the New Census Data. National Community Development Association Palm Springs, CA June 22, 2012 Arthur R Cresce , Jr. Ph.D. Assistant Division Chief for Housing Characteristics Social, Economic and Housing Statistics Division U.S. Census Bureau arthur.r.cresce.jr@census.gov.

nerys
Télécharger la présentation

Keeping Up with the New Census Data

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Keeping Up with the New Census Data National Community Development Association Palm Springs, CAJune 22, 2012 Arthur R Cresce, Jr. Ph.D.Assistant Division Chief for Housing CharacteristicsSocial, Economic and Housing Statistics DivisionU.S. Census Bureauarthur.r.cresce.jr@census.gov

  2. Overview • Background • Overview of the 2000 Long Form and the ACS • Latest on the ACS • Data Quality and the ACS – focus on several key housing measures • Thoughts about the Future

  3. Background

  4. Introduction Census 2000 Basic demographic and detailed housing, social, and economic data ACS Basic demographic and detailed housing, social, and economic data 2010 Census Basic demographic data

  5. Motivation for Change • Call for more timely data • Need to simplify the basic census headcount and control costs • Challenge of collecting high quality detailed data in a census setting

  6. Simplification of the 2010 Census • Census Bureau has concluded that in order to be successful, it must • Increase the currency of detailed housing and population data but • Decrease the decennial census operational complexity • Excellent summary of this in Director’s testimony see: http://www.census.gov/newsroom/releases/pdf/03062012_groves_statement.pdf

  7. Mail Response

  8. Meeting Data User Needs • ACS products were patterned after Census 2000 long form data products • Acknowledge tradeoffs • Period estimates rather than point in time • Reduced precision • Improved data quality • Improved timeliness

  9. Overview of the 2000 Long Form and the ACS

  10. Comparison of 2000 Long Form and ACS

  11. Comparison of 2000 Long Form and ACS * CATI – Computer Assisted Telephone Interview * CAPI – Computer Assisted Personal Interview

  12. Timeliness and Currency of the Data for the ACS • Key additional benefit of the ACS • Detailed characteristics are released less than 1 year after data collection year • Data for all geographic areas released at one time rather than being released one state or several states at a time • Huge undertaking

  13. Latest on the ACS

  14. Data Releases and Current Data Products • Survey Updates and Improvements • ACS Program Review

  15. ACS Data Sets The ACS annually releases… 1-year estimates (2010) • Areas with populations of 65,000+ • The Nation • Every state, the District of Columbia, and Puerto Rico • Every Congressional District • About 25 percent of all counties 3-year estimates (2008-2010) • Areas with populations of 20,000+ 5-year estimates (2006-2010) • All areas down to the census tract and block group level

  16. NOTE: PRINT EDITION ALSO INCLUDED A 2-PAGE SPREAD INSIDE

  17. ACS Data Releases 2011 and 2012

  18. Important Note about ACS Multi-Year Data Products • Geographic boundaries are those of the final year of the period • Dollar values are adjusted to the final year of the period • Multiyear estimates are controlled to the average of the individual year’s estimates for the period

  19. Survey Updates and Improvements Sample Size Increase • Sample expanded from 2.90 million to 3.54 million addresses per year in June 2011 • Benefit to small areas, such as American Indian Reservations

  20. Survey Updates and Improvements Reallocation of Sample to Improve Small Area Estimates • Objective: improve the reliability of the estimates for small areas • Increased sampling rates for small tracts and governmental units • Slightly decreased sampling rates in larger tracts • Began in January 2011

  21. Survey Updates and Improvements Data Collection • Computer Assisted Personal Interview follow-up for 100% of non-mailable and non-responding addresses in: • American Indian areas with estimated American Indian population All Hawaiian Homelands • All Alaska Native Village Statistical Areas • All areas in remote Alaska (began January 2011) (For data collection purposes, “Remote Alaska” comprises all or parts of 14 boroughs in Alaska) • Increases total field workload by about 27,000 cases per year • Benefits areas with small subpopulations

  22. Survey Updates and ImprovementsPlanned New Data Collection Mode: Internet • Testing Underway • Planned Implementation: January 2013

  23. Survey Updates and Improvements:New and Revised Content • New topic – computer and internet usage • Revised topics: • Income • Food stamps – use term “SNAP” (Supplemental Nutrition Assistance Program) • Veteran status • Example of how ACS can accommodate new data needs

  24. American Community Survey Program Review http://www.census.gov/acs/www/about_the_survey/acs_program_review

  25. Internal ReviewGoals • Ensure ACS strategic goals are sound and align with Census Bureau’s mission • Ensure business processes are effective and flexible • Improve systems and processes used to manage ACS program • Improve the governance and program management processes used for the ACS • Make effective use of new technologies

  26. External ReviewGoals • Identified and engaged stakeholderss • Solicited input on program components from organizations or individuals • Get answers to 4 questions: • Are we communicating effectively? • Are our data products meeting your needs? • Are we using the best survey methods? • Do we have the right research agenda? • Committee on National Statistics (CNSTAT), National Academies of Science (NAS) will compile responses and make recommendations to the Census Bureau

  27. Data Quality and the ACS

  28. Example of Activities to Ensure Data Quality • Mandatory response – helps boost mail response rate • Highly trained, permanent staff • Language support in non-English languages – Spanish paper questionnaire, Spanish CATI/CAPI, interviewers who speak other languages as needed • Extensive testing for new and revised questions • Extensive review of the data – compare with data from previous years and with data from other sources such as the Current Population Survey and the American Housing Survey • Responses can only come from household members rather than proxy responses

  29. Example of Activities to Document Data Quality • Provision of data quality page http://www.census.gov/acs/www/methodology/sample_size_and_data_quality • Provision of studies comparing Census 2000 data with ACS data http://www.census.gov/acs/www/library/by_series/acs_census_2000_comparison/ • Report at following url focused on comparison of physical characteristics of housing http://www.census.gov/acs/www/Downloads/library/2004/2004_Love_01.pdf • Provision of guidance on making comparisons with ACS data from previous years http://www.census.gov/acs/www/guidance_for_data_users/comparing_data/ • Provision of margins of error for all statistics

  30. Data Quality

  31. 2006 Content Test: Example of Effort to Improve Data Quality • 2006 first major content test for 23 different question topics - housing questions tested: • Year Structure Built • Rooms and Bedrooms • Plumbing and Kitchen Facilities and Telephone Availability • Property Value See: http://www.census.gov/acs/www/methodology/2006_report_series/

  32. Persons Per Room (Overcrowding) • Number of persons in housing unit divided by number of rooms • Overcrowded – 1.01 or more persons per room • Census 2000 – 5.7% 2005-2009 ACS – 3.0% • Difference of 2.7 percentage points

  33. Percent of Housing Units with 1.01 or More Persons per Room by Tenure Source: American FactFinder, Census 2000, Table H020; 2005-2009 ACS, Table B25014

  34. Differences in Household Size(Percent Distribution) Source: American FactFinder, Census 2000, Tables H019 and H016; 2005-2009 ACS, Tables B25009 and B25010

  35. Issues with Estimates of Rooms • Difficult to determine • 1990 and Census 2000 Content Reinterview Surveys indicate high inconsistency for owner and renter units • C2SS (Census 2000 Supplementary Survey) – fewer 1 and 2 room units than Census 2000 – percentages in the C2SS were lower than in Census 2000 • Overcrowding measure also lower in C2SS – 4.1% vs 5.7% in Census 2000

  36. Impact of 2006 Content Test for Rooms and Bedrooms Question • Rooms and Bedrooms - improvements • Provide a clearer definition of a room • Make it clear that bedrooms are subset of rooms • Provide definition of an efficiency • Result of Test • Improved underreporting of rooms • Increased number of 1-room (efficiency) units • See: http://www.census.gov/acs/www/Downloads/methodology/content_test/H2_Rooms_Bedrooms.pdf

  37. Rooms/Bedrooms Questions Tested in 2006 Old Question New Question

  38. Differences in Number of Rooms for Occupied Housing Units(Percent Distribution) Numbers in red – reflect use of new rooms/bedrooms questions

  39. Comparison of Overcrowding with the American Housing Survey • HUD study assessing impact of using 2005-2009 ACS data for the CDBG formula http://www.huduser.org/portal/publications/commdevl/cdbg_redis_eff.html • Cited American Housing Survey data indicating overcrowding percent around 2.3 percent during the period 2000-2009, suggesting that the ACS estimate is more accurate than the 2000 Census estimate

  40. Year Structure Built: Pre-1940 Housing • Estimate for 2005-2009 ACS 6% larger than Census 2000 estimate • Based on self reporting • Census 2000 content reinterview tests – moderate inconsistency for both owners and renters but level of inconsistency much higher among renters • Conversions may explain part of the difference, but more likely due to differences in self reporting • 2007 study comparing 2000 data with 1999-2001 ACS estimates for the Bronx – but administrative data was closer to the ACS estimate than the Census 2000 estimate

  41. Thoughts about the Future • Research on differences • Vacancy rates - corresponding differences in estimates of households • Persons per room (Overcrowding measure) • Year built – check with administrative data • Content Reinterview Test in 2012 • Use of 5-year estimates – change may only appear gradually

  42. Bottom Line(s) • We are committed to providing the highest quality data we can on a timely basis • We need your help to encourage people to participate in the ACS – good data quality starts with good cooperation • We are willing to work with you to provide relevant materials to get out the word

  43. Data User Support - Educational Materials and Training http://www.census.gov/acs/www/guidance_for_data_users/handbooks/ Measures of sampling error provided with the estimates Compass handbooks for data users

  44. For More Information • User support for data products: 301-763-1405 or acso.users.support@census.gov • Email updates on ACS Program: register at www.census.gov/acs • Program/data products documentation:www.census.gov/acs • ACS Program Review http://www.census.gov/acs/www/about_the_survey/acs_program_review • My contact information: arthur.r.cresce.jr@census.gov; 301-763-3188

More Related