2015-09741. Submission for OMB Review; Comment Request

The Department of Commerce will submit to the Office of Management and Budget (OMB) for clearance the following proposal for collection of information under the provisions of the Paperwork Reduction Act (44 U.S.C. chapter 35).

Agency: U.S. Census Bureau.

Title: American Community Survey.

OMB Control Number: 0607-XXXX.

Form Number(s): ACS-1, ACS-1(SP), ACS-1(PR), ACS-1(PR)SP, ACS-1(GQ), ACS-1(PR)(GQ), GQFQ, ACS CATI (HU), ACS CAPI (HU), ACS RI (HU), and AGQ QI, AGQ RI.

Type of Request: Regular Submission.

Number of Respondents: 3,760,000.

Average Hours Per Response: 40 minutes for the average household questionnaire.

Burden Hours: The estimate is an annual average of 2,455,868 burden hours.

Needs and Uses: The U.S. Census Bureau requests authorization from the Office of Management and Budget (OMB) for revisions to the American Community Survey (ACS). The content of the proposed 2016 ACS questionnaire and data collection instruments for both Housing Unit and Group Quarters operations reflect changes to content and instructions that were proposed as a result of the 2014 ACS Content Review.

The American Community Survey (ACS) is one of the Department of Commerce's most valuable data products, used extensively by businesses, non-governmental organizations (NGOs), local governments, and many federal agencies. In conducting this survey, the Census Bureau's top priority is respecting the time and privacy of the people providing information while preserving its value to the public. The 2016 survey content changes are the initial step in a multi-faceted approach to reducing respondent burden. The Census Bureau is currently carrying out this program of research, which includes several components as discussed briefly below.

One of the areas with strong potential to reduce respondent burden is to reuse information already supplied to the federal government in lieu of directly collecting it again through particular questions on the ACS. The Census Bureau is conducting groundbreaking work aimed at understanding the extent to which existing government data can reduce redundancy and improve efficiency. The tests we are conducting in the next two years will tell us whether existing government records can provide substitute data for households that have not responded to the ACS.

In addition, we continue to look into the possibility of asking questions less often beginning initial efforts on the martial history series of questions. For example, asking a question every other year, every third year, or asking a question of a subset of the respondents each year. We also want to examine ways we can better phrase our questions to reduce respondent concern, especially for those who may be sensitive to providing information.

The outcome of these future steps will be a more efficient survey that minimizes respondent burden while continuing to provide quality data products for the nation. We expect to make great progress during fiscal 2015 on this front, and will be reporting our progress to the Secretary of Commerce at the end of the fiscal year.

Since the founding of the nation, the U.S. Census has mediated between the demands of a growing country for information about its economy and people, and the people's privacy and respondent burden. Beginning with the 1810 Census, Congress added questions to support a range of public concerns and uses, and over the course of a century questions were added about agriculture, industry, and commerce, as well as occupation, ancestry, marital status, disabilities, and other topics. In 1940, the U.S. Census Bureau introduced the long form and since then only the more detailed questions were asked of a sample of the public.

The ACS, launched in 2005, is the current embodiment of the long form of the census, and is asked each year of a sample of the U.S. population in order to provide current data needed more often than once every ten years. In December of 2010, five years after its launch, the ACS program accomplished its primary objective with the release of its first set of estimates for every area of the United States. The Census Bureau concluded it was an appropriate time to conduct a comprehensive assessment of the ACS program. This program assessment focused on strengthening programmatic, technical, and methodological aspects of the survey to assure that the Census Bureau conducts the ACS efficiently and effectively.

In August 2012, the OMB and the Census Bureau chartered the Interagency Council on Statistical Policy (ICSP) Subcommittee for the ACS to “provide advice to the Director of the Census Bureau and the Chief Statistician at OMB on how the ACS can best fulfill its role in the portfolio of Federal household surveys and provide the most useful information with the least amount of burden.” The Subcommittee charter also states that the Subcommittee would be expected to “conduct regular, periodic reviews of the ACS content . . . designed to ensure that there is clear and specific authority and justification for each question to be on the ACS, the ACS is the appropriate vehicle for collecting the information, respondent burden is being minimized, and the quality of the data from ACS is appropriate for its intended use.”

The formation of the ICSP Subcommittee on the ACS and the aforementioned assessment of the ACS program also provided an opportunity to examine and confirm the value of each question on the ACS, which resulted in the 2014 ACS Content Review. This review, which was an initial step in a multi-faceted approach of a much larger content review process, included examination of all 72 questions contained on the 2014 ACS questionnaire, including 24 housing-related questions and 48 person-related questions.

The Census Bureau proposed the two analysis factors—benefit as defined by the level of usefulness and cost as defined by the level of respondent burden or difficulty in obtaining the data, which was accepted by the ICSP Subcommittee. Based on a methodology pre-defined by the Census Bureau with the input and concurrence of the ICSP Subcommittee on the ACS, each question received a total number of points between 0 and 100 based on its benefits, and 0 and 100 points based on its costs. These points were then used as the basis for creating four categories: High Benefit and Low Cost; High Benefit and High Cost; Low Benefit and Low Cost; or Low Benefit and High Cost. For this analysis, any question that was designated as either Low Benefit and Low Cost or Low Benefit and High Cost and was NOT designated as Mandatory (i.e., statutory) by the Department of Commerce Office of General Counsel (OGC) or NOT Required (i.e., regulatory) with a sub-state use, was identified as a potential candidate for removal. The Department of Commerce OGC worked with its counterparts across the federal government to determine mandatory, required, or programmatic status, as defined below:

• Mandatory—a federal law explicitly calls for use of decennial census or ACS data on that question
• Required—a federal law (or implementing regulation) explicitly requires the use of data and the decennial census or the ACS is the historical source; or the data are needed for case law requirements imposed by the U.S. federal court system
• Programmatic—the data are needed for program planning, implementation, or evaluation and there is no explicit mandate or requirement.

Based on the analysis, the following questions were initially proposed for removal:

• Housing Question No. 6—Business/Medical Office on Property
• Person Question No. 12—Undergraduate Field of Degree
• Person Question No. 21—(In the Past 12 mos, did this person) Get Married, Widowed, Divorced
• Person Question No. 22—Times Married
• Person Question No. 23—Year Last Married

For reports that provide a full description of the overall 2014 ACS Content Review methods and results, see “Final Report—American Community Survey FY14 Content Review Results” (Attachment V); additional reports about the 2014 ACS Content Review are also available at http://www.census.gov/​acs/​www/​about_​the_​survey/​methods_​and_​results_​report/​.

Regarding the business/medical office on property question, the Census Bureau received 41 comments from researchers, and individuals. Most of these comments came from researchers who felt that the Census Bureau should keep all of the proposed questions in order to keep the survey content consistent over time, or felt that modifications to the question could potentially make it more useful. Housing Question No. 6—Business/Medical Office on Property is currently not published by the Census Bureau in any data tables. The only known use of the question is to produce a variable for the Public Use Microdata Sample (PUMS), a recode for the Specified Owner (SVAL) variable that allows users to compare other datasets. The Content Review did not reveal any uses by federal agencies, and the comments to the Federal Register notice did not reveal any non-federal uses. Additionally, there were no uses uncovered in meetings with stakeholders, data user feedback forms, or other methods employed to understand the uses of ACS data. Lastly, independent research conducted on behalf of the Census Bureau did not uncover any further uses. Though the question has a low cost, it has no benefit to federal agencies, the federal statistical system, or the nation. The Census Bureau plans to remove this question, beginning with the 2016 ACS content.

Regarding the field of degree question, the Census Bureau received 625 comments from researchers, professors and administrators at many universities, professional associations that represent science, technology, engineering and mathematics (STEM) careers and industries, members of Congress, the National Science Foundation, and many individuals interested in retaining this question. A number of commenters (92) cited the importance of these estimates for research that analyzes the effect of field of degree choice on economic outcomes, including earnings, education, occupation, industry, and employment. University administrators (37) commented that this information allows for analysis of postsecondary outcomes, and allows them to benchmark their graduates' relative success in different fields as well as to plan degree offerings. While some commenters used the estimates to understand fields such as humanities or philosophy (56), the majority of these comments (125) addressed the value of knowing about the outcomes of people who pursued degrees in science, technology, engineering and mathematics. These commenters felt that knowing more about the people currently earning STEM degrees and the people currently working in STEM fields would enable universities, advocacy groups, and policy makers to encourage more people to pursue STEM careers, and to encourage diversity within STEM careers.

The initial analysis of Person Question No. 12—Undergraduate Field of Degree did not uncover any evidence that the question was Mandatory or Required. However, comments to the Federal Register notice uncovered the existence of a relationship between the Census Bureau and the National Science Foundation, dating back to 1960. Over the course of this established relationship, long-form decennial census data was used as a sampling frame for surveys that provided important information about scientists and engineers. These comments demonstrated that the Field of Degree question on the ACS continues this historical use of decennial long-form and ACS data for this purpose, and makes this process more efficient. Many commenters (58) also cited the necessity of the National Survey of College Graduates (NSCG), and recommended retaining the question because it is needed as a sampling frame for the NSCG. Though commenters theorized that the NSCG might still be able to produce STEM estimates without the ACS, a number of commenters (16) thought that doing so would be very expensive, costing as much as $17 million more (1).

Additionally, many comments also indicated uses of this question to understand the economic outcomes of college graduates at local geographic levels, especially those with STEM degrees. These commenters included professional, academic, congressional, and policy-making stakeholders who expressed concerns that the absence of statistical information about STEM degrees would harm the ability to understand characteristics of small populations attaining STEM degrees. Given the importance of this small population group to the economy, the federal statistical system and the nation, bolstered by the new knowledge of historical precedent brought to light by commenters to the Federal Register notice, the Census Bureau therefore plans to retain this question on the 2016 ACS.

Regarding the marital history questions, the Census Bureau received 1,361 comments from researchers and professors, professional associations that represent marriage and family therapists, the Social Security Administration (SSA), and many individuals interested in retaining these questions. SSA commented that it uses the marital history questions to estimate future populations by marital status as part of the Board of Trustees annual report on the actuarial status (including future income and disbursements) of the Old-Age and Survivors Insurance (OASI) and Disability Insurance (DI) Trust Funds. The Department of Health and Human Services (HHS) also uses these questions to distinguish households in which a grandparent has primary responsibility for a grandchild or grandchildren, as well as to provide family formation and stability measures for the Temporary Assistance for Needy Families (TANF) program.

The focus of the proposed elimination is on the marital history questions only with no change to collection of marital status. Over 400 additional comments to the Federal Register notice cited concerns that the proposed elimination of the marital history questions was an indication of whether the government views information about marriage as somehow less valuable than other ACS question topics that were not proposed for removal. While the Census Bureau had always planned to continue collecting information about the “marital status” for each person in a household (Person Question No. 20) and their relationships to each other (Person Question No. 2), the Census Bureau remains sensitive to these criticisms

More than 100 supporters of retaining the marital history questions mentioned their utility for research into marital status changes over time and they correctly noted that there is currently no other national source of the marital history information. As a result, many commenters felt they would not be able to compare marriage characteristics and patterns with other nations in the same depth that is possible today. Similarly, without these questions, the commenters felt that the analysis of changes in marriage events (especially those due to changing societal values and pressures or policy changes) would be less robust. In particular, comments focused on 6 research areas that would be more difficult to analyze without the marital history questions:

• Family formation and stability (23)
• Patterns/trends of marriage and divorce (168)
• Marital effects on earnings, education and employment (45)
• Marital effects on child wellbeing (6)
• Same-sex marriages, civil unions and partnerships (70)
• New government policy effects on marriage (9)

Because the initial analysis of Person Question Nos. 21-23 on marital history did not uncover any evidence that data from these questions were “Required” for federal use at sub-state geographies, those questions received a lower benefit score than many other ACS questions. However, in deference to the very large number (1,367) of comments received on the Census Bureau proposal to eliminate those questions, the Census Bureau plans to retain those questions on the 2016 ACS.

The Census Bureau takes very seriously respondent concerns and recognizes that the Content Review and the resulting, proposed question changes discussed above are only initial steps to addressing them. The Census Bureau has implemented an extensive action plan on addressing respondent burden and concerns. The work completed, and the comments received, on the 2014 Content Review provide a foundation for ongoing and future efforts to reduce burden and concerns. In addition to the immediate content changes (proposed above), the Census Bureau is also currently testing the language on the survey materials that may cause concern such as reminding people that their responses are required by law. In order to be responsive to these concerns about the prominence of the mandatory message on the envelopes, we are conducting research with a subset of ACS respondents in May 2015. Over the summer, we will work with external methodological experts to test other revisions of the ACS mail materials to check respondent perceptions of the softened references to the mandatory nature of participation in the ACS. The preliminary results of those tests will be available in the fall, and the Census Bureau will make changes to the 2016 ACS mail materials based on those results.

Concurrently we also are identifying additional questions that we may only need to ask intermittently, rather than each month or year. The current ACS sample design asks all of the survey questions from all selected households in order to produce estimates each year for small geographies and small populations. However, during the Content Review we learned about over 300 data needs that federal agencies require to implement their missions. We see several potential opportunities to either include some questions periodically, or ask a smaller subset of ACS respondents in cases where those agencies do not need certain data annually. The Census Bureau plans to engage the federal agencies and external experts on this topic during 2015. In addition, we need to assess the operational and statistical issues associated with alternate designs. The alternate designs will result in a reduction in the number of questions asked of individual households.

We are also conducting research on substituting the direct collection of information with the use of information already provided to the government. It is possible that the Census Bureau could use administrative records from federal and commercial sources in lieu of asking particular questions on the ACS.

Lastly, we are examining our approaches to field collection to reduce the number of in-person contact attempts while preserving data quality. For example, based on research conducted in 2012, we implemented changes in 2013 which led to an estimated reduction of approximately 1.2 million call attempts per year, while sustaining the 97 percent response rate for the survey overall. For the person visit operation, we are researching a reduction in the number of contact attempts. We plan to field test this change in August 2015. If successful we would implement nationwide in spring 2016.

We will continue to look for other opportunities to reduce respondent burden while maintaining survey quality. Taken together, these measures will make a significant impact on reducing respondent burden in the ACS.

Affected Public: Individuals or households.

Frequency: Response to the ACS is on a one-time basis.

Respondent's Obligation: Mandatory.

Legal Authority: Title 13, United States Code, Sections 141, 193, and 221.

This information collection request may be viewed at www.reginfo.gov. Follow the instructions to view the Department of Commerce collections currently under review by OMB.

Written comments and recommendations for the proposed information collection should be sent within 30 days of publication of this notice to [e-mail] or fax to [phone].

Dated: April 22, 2015.

Glenna Mickelson,

Management Analyst, Office of the Chief Information Officer.