Skip Main Navigation Skip To Navigation Content

American Community Survey

Census.govAmerican Community Survey › Guidance for Data Users: Comparing ACS Data2011 › Guidance on Comparing 2006-2008 ACS 3-year and 2009-2011 ACS 3-year
Skip top of page navigation

Guidance on Comparing 2006-2008 ACS 3-year and 2009-2011 ACS 3-year

When comparing the 2009-2011 ACS 3-year estimates with the 2006-2008 ACS 3-year estimates, there are three main issues to consider:

  1. Change in geographic boundaries - The 2006-2008 ACS 3-year estimates used legal boundaries as of January 1, 2008 and statistical geographies used Census 2000 vintage geographies. The 2009-2011 ACS 3-year estimates use legal boundaries as of January 1, 2011 and most statistical areas use 2010 Census vintage geographies. For more information on geographic concepts used in the ACS, visit Geography and the ACS.
  2. Change in population controls - The 2006-2008 ACS 3-year estimates and 2009-2011 ACS 3-year estimates use different Census base years for the population estimates used in the ACS weighting. Estimates of population size are not comparable between 2006-2008 and 2009-2011. Estimates of percent distributions, rates, and ratios should be compared with caution. For more details, visit the ACS Research Note Change in Population Controls [PDF 117K].
  3. Change in the questionnaire or coding - The table below provides details by subject for major questionnaire or coding changes.

Race (02)

The ACS question on race was revised in 2008 to make it consistent with the Census 2010 race question. For more information, see the "questionnaire changes document" [PDF 2.2MB]. The reporting of "Other Asian" groups (e.g., Pakistani), "Other Pacific Islander" groups (e.g., Fijian), and the reporting of "White" responses increased at the national level, and the reporting of "Some Other Race" responses decreased at the national level. The change in estimates from 2007 to 2008 may be due to several factors including questionnaire changes, population controls, and methodological changes to the population estimates, and therefore should be used with caution.

Hispanic Origin (03)

The ACS question on Hispanic origin was revised in 2008 to make it consistent with the Census 2010 Hispanic origin question. For more information, see the "questionnaire changes document" [PDF 2.2MB]. The reporting of specific Hispanic groups (e.g., Colombian, Dominican, Spaniard, etc.) increased at the national level. The change in estimates from 2007 to 2008 may be due to several factors including questionnaire changes, population controls, and methodological changes to the population estimates, and therefore should be used with caution. For more information about the changes in the estimates, visit "Changes to the American Community Survey Between 2007 and 2008 and Their Potential Effect on the Estimates of Hispanic Orgin Type, Nativity, Race, and Language".

Ancestry (04)

Beginning with the 2008 ACS 1-year data products, people who reported Irish-Scotch ancestry were classified under "Other groups", whereas in the 2007 ACS and earlier they were classified as Scotch-Irish. Due to this classification change, the Scotch-Irish category decreased from 2007-2008.

Place of Birth (06)

Modest changes were made in 2009 to the region and country groupings. Specifically, Macau (code 225) is now included under the definition of China, and Jan Mayen (code 121) and Svalbard (code 135) are included under the definition of Norway, please see 2009 ACS Code List.

Residence 1 Year Ago (Migration) (07)

Beginning with the 2008 ACS, the migration section of the questionnaire was expanded to collect street address information for residence 1 year ago. The additional detail improves the accuracy of coding and updating residence 1 year ago to current geographical boundaries, which may affect the estimates for movers within the same city/town and movers from principal cities. Areas with recent boundary changes, vague boundaries, or duplicate names are most likely affected. For more information see the report titled Report P.3: Evaluation Report Covering Residence 1 Year Ago (Migration).

Relationship to Householder (09)

The "Child" and "In-law" categories were expanded in the 2008 ACS questionnaire. Use caution when comparing these categories. See "Changes in the Relationship Categories Between 2007 and 2008" for more information on category changes and comparisons. Changes in the questionnaire format and data capture procedures between 2007 and 2008 have resulted in changes in the total number of same-sex unmarried partners in 2008. For more information about same-sex unmarried partners, visit "Changes to the American Community Survey Between 2007 and 2008 and Their Potential Effect on the Estimates of Same-Sex Couple Households".

Household/Family Type (11)

Changes in the questionnaire format and data capture procedures between the 2007 and 2008 ACS have resulted in changes in the number of reported same-sex spouses and hence the total number of same-sex unmarried partners in 2008. All comparisons should be made with caution due to these formatting and data capture changes. For more information about same-sex unmarried partners, visit "Changes to the American Community Survey Between 2007 and 2008 and Their Potential Effect on the Estimates of Same-Sex Couple Households".

Subfamilies (11)

The 2009 edit specifications incorporated changes to improve the identification of married-couple subfamilies with children and single-parent subfamilies. The 2009 data show an increase in these family types.

Educational Attainment (15)

New questions were added to the 2008 ACS CATI/CAPI instrument. Respondents who received a high school diploma, GED or equivalent were also asked if they had completed any college credit. Therefore, data users may notice a decrease in the number of high school graduates relative to previous years because those people are now being captured in the "Some college credit, but less than 1 year of college credit" or "1 or more years of college credit, no degree" categories. For more information see the report titled Report P.2.b: Evaluation Report Covering Educational Attainment [PDF 330KB].

Language Spoken at Home and Ability to Speak English (16)

Changes to collection of data on Hispanic origin may have affected some language characteristics. The observed increase in the native Hispanic population speaking English "very well" is larger than anticipated and should be interpreted with caution. In some cases the overall English language ability in the population may have been affected. See the section on Hispanic Origin (03) above for more information. For more information about the changes in the estimates, visit "Changes to the American Community Survey Between 2007 and 2008 and Their Potential Effect on the Estimates of Hispanic Origin Type, Nativity, Race, and Language".

Disability (18)

The 2008 ACS disability questions are different from the 2007 and prior ACS disability questions, thus comparisons cannot be made. For information about definitions of disability in the ACS, Census 2000 and other Census Bureau surveys, visit the Disability website.

Food Stamp Receipt (22)

In 2010, we added an instruction to the food stamp question incorporating the new program name - Supplemental Nutrition Assistance Program (SNAP). The wording of the food stamp question was revised in the 2008 ACS to ask respondents if they received "food stamps or a food stamp benefit card". Data users may see an increase in food stamp recipiency from the 2007 ACS 1-year estimates to the 2008 ACS 1-year estimates as a result of the questionnaire change. For more detailed information, see the "Evaluation Report Covering Receipt of Food Stamps" [PDF 4.7MB].

Employment Status & Weeks Worked (23)

The Census Bureau introduced an improved sequence of labor force questions in the 2008 ACS questionnaire. Accordingly, we recommend using caution when making labor force data comparisons from 2008 or later with data from prior years. For more information on these questions and their evaluation in the 2006 ACS Content Test, see the "Evaluation Report Covering Employment Status" [PDF 1.04MB] and the "Evaluation Report Covering Weeks Worked" [PDF 504KB]. Additional information can also be found on the Labor Force Statistics website.

Industry and Occupation (24)

The 2006 and 2007 ACS industry data are coded using the 2002 North American Industry Classification System (NAICS), while the 2008 and later years of ACS industry data use NAICS 2007 codes. Industry categories that differ between 2002 and 2007 NAICS are aggregated so that the 3 years of data in the 2006-2008 tables are consistent in display and reflect the NAICS 2007 codes.

The Census occupation codes for 2010 and later years are based on the 2010 revision of the Standard Occupational Classification (SOC). To allow for the creation of 2009-2011 tables, occupation data in the 2009-2011 multiyear file was recoded to 2011 Census occupation codes. The Census occupation codes for 2006-2008 tables are based on SOC 2000. Because of the Census occupation code changes, we recommend using caution when comparing 2009-2011 occupation data with 2006-2008 occupation data. For more information on the Census occupation code changes, please visit Industry and Occupation website.

Bedrooms (25)

Caution should be used when comparing data for Bedrooms between the 2007 and 2008 ACS. Changes made to the 2008 ACS Bedrooms question involving the wording as well as the response option resulted in a noticeable increase in the "No bedroom" category and a decrease in the "1 bedroom" to "3 bedrooms" categories. For more detailed information about these changes, see the Bedrooms section of the "Subject Definitions" [PDF 463KB].

Kitchen Facilities (25)

Caution should be used when comparing data for Kitchen Facilities in certain areas between the 2007 and 2008 ACS. Changes made to the 2008 ACS question wording and translation to the Spanish questionnaire resulted in an inconsistency in the data. For more detailed information about these changes, see the Kitchen Facilities section of the "Subject Definitions" [PDF 463KB].

Telephone Service (25)

Caution should be used when comparing data for Telephone Service Availability between the 2007 ACS and the 2008 ACS. Changes made to the 2008 ACS Telephone Service Availability question involving the structure of the question as well as including an instruction to include cell phones resulted in a noticeable increase in respondents answering "yes" to the question. For more detailed information about these changes, see the Telephone Service Available section of the "Subject Definitions" [PDF 463KB].

Rooms and Occupants per Room (25)

Caution should be used when comparing data for Occupants per Room between the 2007 and 2008 ACS. Changes made to the 2008 ACS Rooms question involving the wording as well as the response option resulted in a noticeable increase in the "1 room" category and a decrease in the "2 rooms" to "6 rooms" categories. For more detailed information about these changes, see the Rooms section of the "Subject Definitions" [PDF 463KB].

Plumbing Facilities (25)

Caution should be used when comparing data for Plumbing Facilities in certain areas between the 2007 and 2008 ACS. Changes made to the 2008 ACS question wording and translation to the Spanish questionnaire resulted in an inconsistency in the data. For more detailed information about these changes, see the Plumbing Facilities section of the "Subject Definitions" [PDF 463KB].

Value of Property (25)

Changes made to the value question between the 2007 and 2008 ACS may result in an inconsistency in the value distribution for some areas. In 2008, the response option for the value question was a write-in. In 2007 and previous years, the value question included categorical response options with a write-in for values over $250,000. The presentation of the data is consistent between 2007 and 2008. For more information about this questionnaire change, see the "Evaluation Report Covering Property Value" [PDF 4.9MB].

Group Quarters (26)

There was a change to the ACS group quarters (GQ) estimation methodology beginning in 2011 to improve sub-state estimates for the total GQ population. This improvement is evidenced in the 2009-2011 ACS data. Due to the change to the ACS group quarters estimation methodology, the 2006-2008 ACS group quarters data should be compared with caution to the 2009-2011 ACS group quarters data. Large fluctuations may be observed at the sub-state level when comparing 2006-2008 to 2009-2011 ACS estimates for the total GQ population.

Health Insurance (27)

The question was not asked in 2007 ACS. For more information about health insurance questions in the ACS and other Census Bureau surveys, visit the Health Insurance website.


[PDF] or PDF denotes a file in Adobe’s Portable Document Format. To view the file, you will need the Adobe® Reader® Off Site available free from Adobe. This symbol Off Site indicates a link to a non-government web site. Our linking to these sites does not constitute an endorsement of any products, services or the information found on them. Once you link to another site you are subject to the policies of the new site.


Source: U.S. Census Bureau | American Community Survey Office | Email ACS | Last Revised: December 04, 2014
Give us your feedback on this website!