ACS NSC - SPD 15 Race and Ethnicity Implementation Plan

OMB Control No. 0607-0810

Request for Non-substantive Change to the

American Community Survey

2024 Statistical Policy Directive No. 15 Implementation Plan

OMB Control No. 0607-0810

U.S. Department of Commerce

U.S. Census Bureau

Purpose

The purpose of this non-substantive change request (NSCR) is to provide a written plan and timeline for the American Community Survey (ACS) and Puerto Rico Community Survey (PRCS) to implement the 2024 Statistical Policy Directive No. 15 (SPD 15) Standards for Maintaining, Collecting, and Presenting Federal Data on Race and Ethnicity.¹

Background

The ACS collects race and ethnicity data from respondents according to the standards outlined by the U.S. Office of Management and Budget (OMB) in SPD 15. In March 2024, OMB issued updates to SPD 15 that must be implemented into all Federal information collections that collect data on race and ethnicity as soon as possible but no later than March 28, 2029.

As indicated in the Census Bureau’s information collection request (ICR) for the 2025 ACS, the race and ethnicity standards will not be updated in data year 2025. As part of the terms of clearance for the 2025 ACS, the OMB requested that the Census Bureau publish a Federal Register Notice (FRN) seeking public input on the impact of different compliance timelines for the ACS, summarize the comments, and provide a written report describing the steps needed for the ACS to comply with SPD 15. As a culmination of the information gathered from these steps, the Census Bureau is submitting this NSCR with a plan and timeline for the ACS to comply with the updated standard.

Summary of Comments

In response to the terms of clearance, an FRN was posted on July 12, 2024, soliciting feedback about the impact the SPD 15 update will have on data users, researchers, and community organizations if it is implemented in either the 2026 ACS or the 2027 ACS.

Through the Interagency Committee on Statistical Policy (ICSP), comments from federal agencies were also solicited given that the ACS serves as a data source for numerous federal statistical agencies and other executive branch agencies.

The Census Bureau received 279 public comments and comments from 25 federal agencies. A summary of the comments received, and the Census Bureau’s response are documented in Attachment A.

Census Bureau Assessment

As a foundational item on the ACS, it is critical that race and ethnicity data produced to meet the updated SPD 15 are accurate. Therefore, we must carefully assess our ability to implement the updated SPD 15 in the ACS while maintaining the data quality standards the public expects.

Local, state, tribal, and federal programs use data on race and ethnicity for a myriad of purposes. Race data are used in planning and funding government programs that provide funds or services for specific groups. These data are also used to evaluate government programs and policies to ensure that they fairly and equitably serve the needs of all racial groups and to monitor compliance with antidiscrimination laws, regulations, and policies.

Staff and leadership from the Census Bureau’s ACS Program have assessed the impact of implementing SPD 15 on ACS systems and operations. Nearly every step in the ACS survey life cycle, from content development through data dissemination, is impacted by this change. While prior content testing provides a solid basis for the questions and associated variables for data collection instruments, the coding, editing, imputation, and data products require additional time for development. That development time also includes outreach to the public and key stakeholders about the SPD 15 changes.

This assessment, along with the proposed implementation timeline, is documented in Attachment B.

Final Recommendation

Based on the Census Bureau’s assessment, feedback from the public, along with feedback from other federal agencies that rely on ACS data, the Census Bureau proposes to begin collecting race and ethnicity data under the new standard beginning with the 2027 ACS.

Attachments

Attachment A. Summary of Comments on ACS Implementation of SPD 15

Attachment B. Implementing Updated SPD 15 in the ACS

Attachment A.

Summary of Comments on ACS Implementation of SPD 15

On July 12, 2024, the Census Bureau published a Federal Register Notice (FRN) seeking public input on the impact of different timelines for the American Community Survey (ACS) to comply with SPD 15. Through the Interagency Committee on Statistical Policy (ICSP), comments from federal agencies were also solicited given that the ACS serves as a data source for numerous federal statistical agencies and other executive branch agencies. This document provides a summary of the comments received and the Census Bureau’s response to the comments.

1. Public Comments

Table 1 summarizes the comments received on the July 2024 FRN: ACS Timeline for Implementing Updated 2024 Race and Ethnicity Data Standards.

Table 1. Public Implementation Timeline Feedback by Comment Topic

TOTAL PUBLIC COMMENTS		279
Comment on SPD 15 Policy		231
Discussed change to Black category	123
Discussed change to MENA category and need for more research	108
Did not comment on timing		8
Recommended postponing or not implementing		12
Support 2027		6
Support earliest year after research		5
Support 2026		17

• Most of the comments received (over 80%) were comments on the SPD 15 policy itself.

  • 123 comments related to requests to change the “Black or African American” category to distinguish people descended from slaves vs. immigrants from Africa or the Caribbean. Various requests were made about the language of the categories or how to address this issue. A majority of these comments did not comment on the timeline for implementing the current SPD 15. Five comments explicitly stated the updated standard should not be implemented until this issue is addressed.

The U.S. Census Bureau collects race and ethnicity data following standards set by the U.S. Office of Management and Budget (OMB). Per the guidance in OMB’s 2024 Statistical Policy Directive No. 15, the Census Bureau will continue to use the category name “Black or African American” for the ACS.

• 108 comments related to requests to include Armenian in the MENA classification. Most of these commenters supported 2027 implementation in order to allow sufficient time to adjust the classification and engage the community in discussions. However, comments also stated that if the Census Bureau decides not to pursue research prior to the 2027 ACS implementation, it should proceed as quickly as possible with implementation in 2026 in order for this necessary research to be conducted before implementation on the 2030 Census.

The Census Bureau plans to implement the updated 2024 standards in the ACS as directed by OMB. The updates to SPD 15 were informed by robust empirical research; extensive engagement with experts, scholars, organization leaders and communities across the country; and the successful and meaningful collaboration among federal agency leaders and experts on the Interagency Technical Working Group on Race and Ethnicity Standards. In its decision, OMB noted that although several commenters expressed interest in explicitly including Armenian as part of the MENA category, the 2015 National Census Test found that most respondents who identify as Armenian did not select MENA when it was offered. The OMB advised that additional Federal agency research is needed on these groups to monitor their preferred identification.

The Census Bureau is not pursuing empirical research to change the race and ethnicity question design at this time. The OMB has established a Federal Committee on SPD 15 that will manage a government-wide research agenda and undertake regular reviews of the standards. Any plans for experimental testing will be considered at a later time and in collaboration with OMB’s Federal Committee on SPD 15. Once the Census Bureau has implemented the updated standards with a combined race and ethnicity question, a dedicated MENA category, multiple detailed checkboxes and examples, and dedicated write-in response areas, we will have the ability to examine how respondents from all communities are self-identifying within the framework of the updated race and ethnicity standards.

Additionally, the Census Bureau will be seeking public feedback through a FRN later this year on how detailed race and ethnicity populations will be coded in the ACS. The Census Bureau will consider all feedback provided in response to the FRN, but all updates to how groups are coded must also be supported by strong research and evidence and align with the definitions of the minimum reporting categories in the updated SPD 15 before making any changes.

• Eight comments did not mention the timeline and are out of scope for this document. They raised various issues:

• One supported implementation but did not comment on the timeline.

• Two indicated that race is a false identity.

  • One raised concerns and questions about the crosswalk needed for the U.S. Department of Veterans Affairs purposes.

• One suggested that the race question should be changed to ask about visual identification but keep categories as is.

• One suggested race categories should be customized based on region of the respondent to allow for easier self-identification; bridging factors should also be sub-national.

• One provided their race without further comment.

• One indicated comments were in an attachment, but no attachment was provided.

• Twelve comments suggested postponing implementation or that the Census Bureau should not implement SPD 15. Reasons for postposing or not implementing included requesting a change to the Black or African American category (5) as discussed above, concerns about data breaches and safety (4), waiting to implement the updates in the ACS until 2030, when they are also implemented into the decennial census (1), concerns that the justification for collection and use of race data is not clear (1), and needing to change the question wording and categories, and conduct testing before implementing (1).

The Census Bureau uses confidential survey answers to create statistics, and no one is able to identify individuals’ survey answers from the statistics we produce. The Census Bureau is legally bound to strict confidentiality requirements. Individual records are not shared with anyone, including federal agencies and law enforcement entities. Data on race and ethnicity are used by local, state, tribal, and federal programs, and they are critical factors in the basic research behind numerous policies, particularly for civil rights. Race and ethnicity data are used in planning and funding government programs that provide funds or services for specific groups.

The Census Bureau must implement the updated SPD 15 as expeditiously as possible, and within OMB’s required 5-year implementation period (no later than March 28, 2029). The updates to SPD 15 were informed by robust empirical research; extensive engagement with experts, scholars, organization leaders and communities across the country; and the successful and meaningful collaboration among federal agency leaders and experts on the Interagency Technical Working Group on Race and Ethnicity Standards. Many of the SPD 15 updates are supported by Census Bureau research findings over the past decade. Any plans for experimental testing will be considered at a later time and in collaboration with OMB’s Federal Committee on SPD 15.

• Six comments supported 2027 implementation; they were primarily concerned with accuracy of the data and expressed that if the Census Bureau needs another year to ensure quality data, they were supportive of that. Commenters also expressed a desire to be involved in discussions about coding, bridging, and data products.

• Five comments supported either year for implementation, with most requesting implementation as soon as feasible after completing research and stakeholder outreach. The request for research varied from wording and category research to bridging and data products. These commenters also indicated that if the Census Bureau was not planning to do any additional research, then the updates should be implemented in 2026.

• Of the seventeen comments that supported implementation in 2026, most referenced the need for the low-level data sooner for specific sub-groups and to address inaccurate reporting and burden for respondents. They also mentioned the need for data users to have data ahead of 2030, both to prepare for 2030 data and to provide an opportunity to correct issues prior to 2030 data products.

The Census Bureau is appreciative of the publics comments on the timeline to implement SPD 15 in the ACS. In making a final recommendation about the timeline the Census Bureau weighed the need for producing data under the new standard as quickly as possible against the need for accurate data.

2. Federal Agency Comments

Through the Interagency Committee on Statistical Policy (ICSP), 25 federal agencies responded to a request to identify impacts to their agency and programs if ACS implemented the updated SPD 15 in 2027 as opposed to 2026. Table 2 summarizes the comments.

Table 2. Federal Agency Feedback

TOTAL FEDERAL AGENCY RESPONSES		25
Support for 2027 Implementation		20
Minimally impacted, not impacted, or have no concerns about timing of a 2027 implementation	12
Support minimizing risks and ensure data quality	2
Aligns with agency and/or survey plans	2
No concern with timing of implementation	4*
Expressed some concern		5

Twenty of these agencies indicated there was no impact to their agency or programs of the timing of ACS implementation. However, four agencies stressed the value of communication & coordination on data collection, coding, editing, bridging factors, imputation, tabulation, and other methodological issues. (Note, in addition to these four agencies, three agencies that expressed concern about the timeline also indicated the importance of communication and coordination.)

Five agencies expressed some concerns about the timing of ACS implementation. Census Bureau staff met with these five agencies (plus the Centers for Medicare & Medicaid Services) to discuss the use of ACS for these surveys and programs in more detail. A summary of the agency, their concern, and discussion is below.

Table 3. Concerns Raised by Agency and Status

Agency	Concern	Status
National Center for Health Statistics (NCHS)	Expect Vital Statistics to take long to transition. Primarily concerned with comparability of data.	This concern exists regardless of whether ACS implements in 2026 or 2027. Census Bureau staff can work with NCHS on the bridging algorithm.
Centers for Disease Control and Prevention (CDC)	Use ACS for weighting and population denominators for various programs.	Staff at CDC were open to using Population Estimates Program (PEP) or bridging factors, with Census Bureau support. Census Bureau staff will continue discussions with CDC about options.
Health Resources and Services Administration (HRSA)	National Survey of Children’s Health (NSCH) uses ACS for raking: age, sex, race, household poverty ratio, household size, education.	NSCH uses prior year ACS data which would mean use for NSCH in 2027 or 2028. Theoretically PEP could be used for race or bridging factors could be used but it would be a tradeoff of effort vs implementation timeline.
HRSA	National Sample Survey of Registered Nurses (NSSRN) uses ACS for weighting; cannot use PEP because of the subpopulation of interest.	NSSRN is conducted every four years. The next iteration of the survey is in 2026, which will use 2025 ACS data for weighting. The 2025 ACS will not have updated race data. The 2030 iteration will use 2029 ACS data, which will have the updated standards applied. NSSRN is also considering conducting the survey every two years. If the survey is conducted in 2028, data from the 2027 ACS with the updated standard will be available for use. Therefore, there is no impact of the timing of ACS implementation on NSSRN.
National Center for Science and Engineering Statistics (NCSES)	The National Training, Education, and Workforce Survey (NTEWS) uses ACS as the frame and the sample is stratified on sex, race, educational attainment, and Science Technology Engineering and Math (STEM) occupations.	If NTEWS stays on the current schedule it delays use of ACS updated race data for sampling by two years. PEP cannot be used. This impact may be mitigated if NTEWS is combined with National Survey of College Graduates (NSCG) and implemented on the NSCG schedule.
NCSES	Analytical and congressionally mandated STEM reports rely on ACS public use microdata sample (PUMS) data. Reports use 3-year-old ACS data.	If ACS implements in 2026, the reports released in 2029 would use the new standard. If 2027, then reports would be updated in 2030.
Internal Revenue Service (IRS)	Use of ACS data for Bayesian Improved First Name and Surname Geocoding method of imputing race and ethnicity data when not collected (reported).	The 5-year ACS data will be available beginning in December 2028 and use 1 year of race and ethnicity data collected under the new standard and 4 years worth of race and ethnicity data cross-walked from the old standard to the new standard. IRS is evaluating a pilot of noise injected race and ethnicity microdata. Either approach (ACS data or the microdata) will require a crosswalk and bridging of race and ethnicity data under the current standard to the new standard.

The Census Bureau expects to have continued conversations with these agencies. The Census Bureau will also participate on interagency subgroups to discuss methodological best practices related to implementation.

Attachment B.

Implementing Updated SPD 15 in the ACS

TABLE OF CONTENTS

1. INTRODUCTION 12

2. ACS SURVEY LIFECYCLE OVERVIEW 12

3. FRAME & SAMPLE 13

4. DATA COLLECTION AND CAPTURE 14

4.1 Respondent Contact Strategy and Materials 15

4.2 Instrument Development 15

4.3 Paper Data Capture 16

4.4 Training 16

4.5 Case Management and Control System 17

4.6 Language Support 17

5. POST-DATA COLLECTION PROCESSING 18

5.1 Data Integration 18

5.2 Write-in Response Coding 19

5.3 Edit Input 20

5.4 Editing, Imputation, and Review 21

5.5 Disclosure Avoidance 22

5.6 Weighting and Estimation 23

6. DATA PRODUCTS 23

6.1 Data Products Development 24

6.2 Data Products Review 26

6.3 Data Reasonableness Review 26

6.4 Data Products Dissemination 27

7. COMMUNICATIONS 27

8. IMPLEMENTATION DEPENDENCIES AND CONSTRAINTS 27

8.1 County-Level Bridging Factors 27

8.2 Updated Population Estimates 28

8.3 Race/Ethnicity Coding Improvement Project 28

8.4 Survey and Enterprise Innovations and Initiatives 29

8.5 2025 Content Changes 29

8.6 Next ACS Content Test 30

8.7 Disability Stakeholder Engagement Activities 31

8.8 Sexual Orientation and Gender Identity Content 32

9. CONCLUSIONS 32

10. REFERENCES 34

Appendix A. Preliminary SPD 15 Implementation Schedule for ACS 35

1. INTRODUCTION

On March 28, 2024, the Office of Management and Budget (OMB) published a set of revisions to Statistical Policy Directive No. 15 (SPD 15): Standards for Maintaining, Collecting, and Presenting Federal Data on Race and Ethnicity (OMB, 2024). This is the first revision of SPD 15 since 1997. Once OMB updated SPD 15, the Census Bureau began exploring options for implementing the new standards in the American Community Survey (ACS).² OMB requires an implementation plan be submitted within 18 months and full implementation by March 29, 2029.

The Census Bureau is responsible for producing the best possible data to provide accurate information about America’s people, places, and economy. This is particularly important for our flagship survey, the ACS, which serves as the foundation for federal, state, and local policymaking. Therefore, we must carefully assess our ability to implement the updated SPD 15 in the ACS while maintaining the data quality standards the public expects.

The race and ethnicity questions are foundational items on the ACS. As foundational items, the data are used for edits for other topics as well as cross-tabulations in many data tables. Therefore, it is a complex undertaking to implement changes to how race and ethnicity data are collected and produced. Adding to this complexity, the post processing for this change will be done as the Census Bureau is transitioning to a modern platform for data collection using cloud-native software and open-source technologies to process and analyze data.

ACS Program staff and leadership have fully assessed the impact on ACS systems and operations of implementing SPD 15 in the ACS. The team evaluated the value and priority of the ACS project portfolio, focusing on major program transformations and the implementation of SPD 15 while taking into consideration regular ongoing ACS production work. This holistic assessment approach is essential for the ACS to continue achieving its mission while adapting to change. This report documents the steps needed to comply with the updated SPD 15 for the ACS survey lifecycle operations and systems. It also discusses competing priorities that impact resources needed to implement the changes for the ACS.

2. ACS SURVEY LIFECYCLE OVERVIEW

The implementation timeline of SPD 15 was evaluated across the ACS survey lifecycle. The Census Bureau categorizes survey work into functional areas that comprise the day-to-day production and operational work required for a survey like the ACS. Five major components of the ACS survey lifecycle that are impacted by the revision to SPD 15 are as follows:

Each component of the survey lifecycle includes many operations and systems that make up the ACS survey. This document assumes the reader has a basic knowledge of the ACS. For more information about the design of the American Community Survey and details of the full set of methods and procedures, see the ACS and PRCS Design & Methodology Report (U.S. Census Bureau, 2022).

Implementing the updated SPD 15 requires significant changes to data collection, post-data collection processing, and data product systems. To understand the impact of these changes, Census Bureau staff held a series of meetings to assess what would need to be done, how it would be conducted, and the resources needed for the implementation of SPD 15.

• Several meetings were conducted to evaluate each component of the survey lifecycle, what changes would be needed, the level of effort to make the changes, and when information would be needed from subject matter experts to implement the change under various scenarios.

• Several all-day meetings were held to review major projects and initiatives to determine how they align with the strategic vision of the ACS and Census Bureau. Staff assessed the resource impact and associated risks of implementing SPD 15 concurrent with each of the top priorities for the program.

This assessment assumes no additional content testing is needed before implementation of the new combined race/ethnicity question on the survey.

3. FRAME & SAMPLE

The ACS and PRCS each consist of two separate samples: housing unit (HU) addresses and residents of group quarters (GQ) facilities. We derive the sampling frames from which we draw these samples from the Census Bureau’s Master Address File (MAF) Frame. The MAF contains mailing and location address information, geocodes, and other attribute information about each living quarter. A geocoded address is one for which state, county, census tract, and block have been identified.

The updated SPD 15 race and ethnicity standard has no impact on the frame or sample selection, with one exception noted in the table below.

System/Operation	Activities
HU Frame Creation	No changes needed
GQ Frame Creation	No changes needed
Bureau of Prisons GQ Frame Creation	No changes needed
HU Sampling	No changes needed
GQ Sampling	No changes are needed at this time. Note: Bureau of Prisons sampling would be impacted when the Bureau of Prisons makes changes to their race and ethnicity data, which serves as an input to the sampling.

4. DATA COLLECTION AND CAPTURE

Work in this area includes data capture methods and procedures for paper questionnaires, training procedures for interviewers, workload management, and language assistance.

The work in Data Collection and Capture is focused on HU and GQ data collection operations. This includes:

• Respondent contact strategies and materials, such as notifications and reminder approaches

• Content development and instrument development and testing for paper questionnaires, internet data collection, Computer-Assisted Personal Interviewing (CAPI), Computer-Assisted Telephone Interviewing (CATI), Telephone Questionnaire Assistance (TQA), Failed Edit Follow-up (FEFU), and Reinterview

• Data capture methods and procedures for paper questionnaires

• Training procedures for interviewers administering the survey over the phone or in person, as well as answering questions and conducting follow-up interviews

• Workload and response management (control system)

• Language support

The implementation of SPD 15 impacts most, but not all, data collection and capture components. Given prior experience with implementing similar content in the 2016 ACS Content Test, the implementation of SPD 15 within the data collection and capture systems can be done with current resources without major risk.

1. Respondent Contact Strategy and Materials

The updated SPD 15 has no impact on the mail contact materials, such as letters and reminders, or on the letters that interviewers have available to them.

2. Instrument Development

Census Bureau staff have decades of experience implementing question changes in the ACS data collection instruments. Proposed content changes follow a change request process that includes documentation and committee review along with specification updates, programming, and testing. Updates to content include both English and Spanish for all modes of data collection for the HU and GQ instruments.

System/Operation	Activities
HU and GQ Paper questionnaire	Question wording changes Design new questionnaire Print new questionnaires
HU and GQ Internet Instruments	Question wording, format, and help text changes to align with the mode of administration (i.e., the question is not the same as paper) Specify output variable names and values Program instrument changes (HU internet English, HU internet Spanish, GQ internet English, GQ internet Spanish) Test instruments and verify output
HU and GQ Personal Interview (CAPI Operation)	Question wording, format, and help text changes to align with the mode of administration Specify output variable names and values Program instrument changes (HU CAPI English and Spanish, GQ CAPI, HU CATI English and Spanish, and GQ CATI) Test instruments and verify output
Telephone Questionnaire Assistance	Question and help text changes Specify output variable names and values Test instrument and verify output
Failed-edit Follow-up (FEFU)	Question and help text changes Specify output variable names and values Test instrument and verify output Develop a variable crosswalk Note: There is only one FEFU instrument which is administered based on the date of data collection. Therefore, for example, a case in sample in 2026 but that responds in January of 2027 would be asked the 2027 question version. Implementing this change requires a crosswalk of variables.

3. Paper Data Capture

The USPS returns all completed ACS paper questionnaires to the Census Bureau’s National Processing Center. The paper questionnaires are checked in and sorted by type of questionnaire. They are then scanned, and the scanned images continue through the data capture process. The scanning system uses optical mark recognition (OMR) and optical character recognition (OCR) to capture data from the paper questionnaires. Keyers capture data that the scanning system was unable to interpret with high confidence. When content of the questionnaire changes, the scanning system and keying system need to be updated.

System/Operation	Activities
Paper Questionnaire Check-in	No changes needed
Scanning System	Update requirements, such as question placement, variable names, and values Program system changes Test system and verify output
Keying System	Update requirements, such as variable names, valid values, and keying rules Program system changes Test system and verify output Update training materials and manuals for keyers Train keyers

4. Training

While training materials will need to be updated, we anticipate those updates can be completed with current resources without major risk.

System/Operation	Activities
HU and GQ Personal Interview (CAPI Operation)	Update training materials and manuals Train interviewers
Telephone Questionnaire Assistance	Update training materials and manuals Train interviewers
Failed-edit Follow-up	Update training materials and manuals Train interviewers

5. Case Management and Control System

The ACS case management and control system includes detailed information documenting operational outcomes for every ACS sample case. It is used to manage the operational flow of cases through the various data collection phases daily. There are no changes to the control file needed because of revisions to SPD 15.

Data from each mode of data collection are also processed in this step. Every day data are received from the various systems used to collect and capture data from respondents in both HUs and GQs. Changes are needed to the system that receives this data based on the revisions to SPD 15. A new variable glossary for the race and ethnicity question is required to define new variable names and ranges of values for each mode of data collection.

The resources and risk regarding system changes is discussed in more detail in sections 5 and 6.

System/Operation	Activities
ACS Case Management and Control System	No changes Needed
Daily Processing System	Requirements and specification updates to the variable glossary (variable names, range of values) by mode Program system changes Test system changes

6. Language Support

All revised content in the data collection instruments, help text, and other materials needs to be translated into Spanish. We anticipate that translation can be completed with current resources without major risk.

System/Operation	Activities
All data collection instruments	Translate question text and help text
Training materials and manuals for field representatives	Translate question text and instructions

5. POST-DATA COLLECTION PROCESSING

The collected data undergo several processing steps before they are ready for use in data products. Data preparation and processing are critical steps in the survey process, particularly in terms of improving data quality. Developers of large ongoing surveys, such as the ACS, develop and follow rigorous procedures and rules to guide these processes and ensure they are done accurately and consistently.

The primary purpose of data preparation and processing is to take the response data gathered from each survey collection mode to the point where they can be used to produce survey estimates. Data returning from the field typically arrive in various stages of completion, from a completed interview with no problems to one with most or all the data items left blank. There can be inconsistencies within the interviews, such that one response contradicts another. There may also be multiple interviews returned from the same household. These returns could contain different answers to the same question.

The implementation of SPD 15 will have significant impacts on the post-processing activities of the ACS. Many changes will require more time than usual to update requirements, program the changes, and test the changes. For example, changes to the race and ethnicity questions in the 2020 ACS took over 6 months to program and test, which was only possible because technical staff utilized work completed previously for the 2020 Census. Several staff will need to reprioritize other projects to complete this work, putting the deferred work at risk and creating both work and cost inefficiencies. Technical staff responsible for most of the systems discussed in this section are also responsible for an ongoing massive transformation of the ACS systems to implement cloud-based services and open-source software. That transformation is on a conflicting timeline with the SPD 15 changes due to upcoming decommissioning of server-based platforms and transition to new data collection systems. More information on these initiatives can be found in section 8.4.

The risks to post processing by rushing the implementation of SPD 15 are significant. An error identified after the release of products would involve large and unplanned costs, time delays, and call in into question research and policies based on flawed data. It could also contribute to the mistrust of, and reduced confidence in, the Census Bureau.

1. Data Integration

In Section 4.5, we discussed changes needed to process data from each mode of data collection. In this step, daily responses from different modes are “normalized”, collated, and accumulated, forming Data Collation Process files (DCP files). In this context, “normalized” means that the data are aligned to form one set of variables and values.

For race and ethnicity, this involves aligning data across the modes to account for different administrations of the question. In the paper mode, the race and ethnicity question will be presented as one question; in automated modes, the question will be “waterfalled” where an initial question collects the major race and ethnicity categories, and based on selections, subsequent questions are presented to collect further detail. Additionally, write-in responses are initially evaluated as blank, do not know, refused, or requiring coding.

System/Operation	Activities
Data Collation Process	Update requirements. The same inputs and requirements from data processing are needed for data integration. Develop a crosswalk to map the data from the old content to the new content and from the new content to the old content. Update software specifications Program system changes Test system changes

2. Write-in Response Coding

The write-in entries from the DCP files are then ready to be coded as part of monthly coding operations.

The Census Bureau is in the process of conducting its Race/Ethnicity Coding Research Project. A key component of this project is seeking feedback from the public through an FRN that will be published in Fall 2024.

Any updates to the code list that result from the Race/Ethnicity Coding Research Project will require updates to the code list, coding Masterfile, crosswalk, edit specifications, data product specifications, etc.

System/Operation	Activities
Automated Coding System	Update requirements (e.g., variable names) Update the code list and Masterfile with results of the Race/Ethnicity Coding Improvement Project Update software specifications Program system changes Deliver coding Masterfile for production Test system changes
Clerical Coding System	Update requirements and code list like the automated coding system Program system changes Test system changes Update training materials Conduct coder training Update workload estimates Update staffing plans

3. Edit Input

Once all response data for the year have been processed, a series of steps are taken to produce Edit Input Files. These are created by merging operational status information (such as whether the unit is vacant, occupied, or nonexistent) for each HU and GQ persons within a facility with the files that include the coded response data. These combined data then undergo several processing steps before they are ready for use in data products. First, the response type and number of people in the household unit are assessed to determine inconsistencies. Second, the return is examined to establish if there are enough data to count the return as complete. Third, if more than one response is received from a sample address, we assess and select which return will be used. The third step is called the Primary Selection Algorithm (PSA). Steps one and two are referred to in the table below as pre-PSA processing.

In general, the edit input processes are less impacted than other processing steps as it essentially passes the data along, except what we use to determine a sufficient partial and to assess which form to use if multiple returns are received. In these cases, variables and values need to be updated.

System/Operation	Activities
Pre-PSA Processing	Update requirements with new variable names and values Program system changes Test system changes
Primary Selection Algorithm (PSA)	Update requirements with new variable names and values Program system changes Test system changes
Post-PSA File Creation	Update requirements with new variable names and values Program system changes Test system changes

4. Editing, Imputation, and Review

The main steps in this process are editing and imputation, generating recoded variables, reviewing edit results, and creating input files for the next steps.

Editing involves resolving inconsistent data. For example, a person reports a bachelor’s degree as their highest level of educational attainment, but indicates they are currently enrolled in High School.

When data are missing, it is standard practice to use a procedure called imputation to fill in missing responses. Filling in missing data provides a complete dataset, making analysis of the data both feasible and less complex for users. Imputation can be defined as the placement of one or more estimated answers into a field of a data record that previously had no data or had incorrect or implausible data (Groves et al., 2004).

New variables are created during data processing. These recoded variables, or recodes, are calculated based on the response data. Recoding usually is done to make commonly used, yet complex, variables user-friendly and to reduce errors that could occur when data users incorrectly recode their own data. For example, creating a variable to represent people 60 and over or creating a poverty-status indicator.

After editing and imputation are complete, Census Bureau subject matter analysts review the resulting data files. Once the subject matter analysts have approved data within the edited files, and their associated recodes, the files are ready to serve as inputs to the weighting operation and disclosure review.

For 5-year data, several steps must be applied to the previous years’ final edited data to make them consistent for multiyear processing. This includes applying updated geography for the residence address, place of work, and migration data, inflating income data, aligning variables and values that have had content changes, etc.

In addition to developing new edits for race and ethnicity, 16 other topics on the ACS use race and ethnicity data in their edit or imputation logic, including Ancestry; Citizenship Status; Computer and Internet; Employment Status; Fertility; Health Insurance Coverage; Income and Earnings; Industry, Occupation, and Class of Worker; Journey to Work; Language Used at Home; Marital Status; Migration; Period of Military Service; Place of Birth; Public Assistance; and School Enrollment.

System/Operation	Activities
Edit and Imputation System for 1-year Data	Updated Requirements and Specifications: Variable names and variable recodes Develop race and ethnicity edit Update other question edits that utilize race and ethnicity Change hot deck imputation methods Program system changes Test system changes
Edit Review Tools	Update variables and edit changes Review tool content for many other edits that use race and ethnicity are impacted
Subject Matter Expert Review for 1-year Data	Update review programs with new variables and values Test the race/ethnicity edit Develop a national-level file to use for initial testing of the updated race/ethnicity edit Conduct subsequent testing using state-level files for topics other than race Additional time for testing the edit is needed beyond the normal testing period due to the complexity of the edit Testing and finalizing edit specifications and programs for race and ethnicity must be done before other subjects’ edits because there are ripple effects on other edits and products
Edit and Imputation System for 5-year Data	Develop variable crosswalk for race and ethnicity data Program system changes Test system changes
Subject Matter Expert Review for 5-year Data	Verify crosswalk of race and ethnicity data

5. Disclosure Avoidance

The ACS disclosure avoidance methodology is based on the swapping, or interchanging, of a small percentage of household records between pairs of households in different geographic areas. The selection process for swapping is highly targeted to affect records with the most disclosure risk. It is not expected that the changes due to SPD 15 will cause a major change in the current methodology.

System/Operation	Activities
Disclosure Protection System	Assess disclosure risk using early versions of the edits Update specifications Program system changes on both the 1-year and 5-year data Test system changes

6. Weighting and Estimation

Changes to the weighting and estimation systems do not require a lot of effort but are dependent on the development of county-level bridging factors (see section 8.1 for more information) being applied to the population estimates produced by the Population Estimates Program (see section 8.2 for more information) and to the 5-year microdata.

System/Operation	Activities
GQ Person Weighting System HU Weighting System Multiyear Estimation System	Assess and develop revised methods to create the population controls for use in the weighting Update specifications to utilize new population controls Program system changes Test system changes

6. DATA PRODUCTS

The activities for Data Products and Data Dissemination include the production, review, and release of the 1-year and 5-year ACS data estimates. ACS data products include the tables, reports, and files that contain estimates of social, economic, housing, and demographic characteristics. These products cover geographic areas within the United States and Puerto Rico.

The actual processing of the data products cannot begin until all response records for a given year or years are edited and imputed in the data preparation and processing phases, disclosure avoidance techniques are applied, and the final weights are determined. Using the weights, the sample data are tabulated for a wide variety of characteristics according to the predetermined content. These tabulations are done for the geographic areas that have a sample size sufficient to support statistically reliable estimates, except for 5-year estimates, which are available for all small geographic areas down to the census tract and block group levels. The Public Use Microdata Sample (PUMS) files are also created which provide a microdata file that is a subset of the full sample data.

After the estimates are produced, Census Bureau subject matter analysts review them for reasonableness. When the estimates have passed the final review, they are released to the public. PUMS data files are reviewed in a separate process.

Like the impacts on post-data collection processing, many changes to data products will require more time than usual to update requirements, program the changes, and test the changes. Technical staff responsible for data products are also responsible for transforming data product systems to implement cloud-based services and open-source software.

1. Data Products Development

The race and ethnicity changes will impact more than 600 tables, which is over one-third of existing 1-year or 5-year ACS tables, for Race and Ethnicity and other topics that are cross-tabulated by Race and Ethnicity, plus the creation of an unknown number of new tables. These modifications will comply with OMB guidance on tabulation procedures for SPD 15. In addition to standard annual ACS tables, multiple ACS special tabulations will be affected such as the Citizen Voting Age Population (CVAP) by Race and Ethnicity and the Selected Population Tables and American Indian and Alaska Native Tables (SPT-AIANT).

System/Operation	Activities
Detailed Tables	Update Specifications Variable name and code changes have a big impact on the specifications. We need a good test plan to verify all references are caught, and that there is internal consistency. Detail specs and downstream impacts, derived products, population groups, iteration specs, table universes, etc. Redefine householder race and ethnicity (this information feeds into several products) Define new tables for Middle Eastern and North African (MENA) populations and update existing tables to combine race and ethnicity Revise universes of existing tables Program system changes The table generator system should not have to change much but is very sensitive to what is in the specifications. Test system changes
Data Profiles	Specification Updates (see notes above) Program system changes Test system changes
Comparison Profiles	Specification Updates (see notes above) Program system changes Test system changes
Geographic Comparison Tables	Specification Updates (see notes above) Program system changes Test system changes
Narrative Profiles	Specification Updates (see notes above) Program system changes Test system changes
Ranking Tables	Specification Updates (see notes above) Program system changes Test system changes
Subject Tables	Specification Updates (see notes above) Program system changes Test system changes
Selected Population Profiles	Specification Updates (see notes above) Program system changes Test system changes
Selected Population Tables and American Indian and Alaska Native Tables	Specification Updates (see notes above). Note that this product is not an annual product. It is produced typically every 5 years. Based on the changes to content, we need to determine when this product will be produced and released next as the current schedule may not be the best for data users. Program system changes Test system changes
Public Use Microdata Sample (PUMS)	Specification Updates (see notes above). We need to redefine race and ethnicity categories on the PUMS to include MENA while still adhering to the disclosure avoidance requirements. Program system changes Test system changes

2. Data Products Review

In data product review, data tables are rendered on a test platform for staff review. Subject matter experts verify that estimates are produced according to the specifications and displayed correctly.

System/Operation	Activities
Data Review System	Update filtering options for race and ethnicity tables Revise population groups hierarchy Program review system changes Test review system changes Update subject matter expert review programs

3. Data Reasonableness Review

In data reasonableness review, subject matter experts review the weighted data to determine whether estimates are consistent with our base of knowledge about the characteristics being estimated, that is, are they reasonable over time and within a geographic area when compared to changes in other estimates.

System/Operation	Activities
Reasonableness Review tools	Create rules for past year processing Update system specifications Program system changes Test system changes
Data Reasonableness Review	Update review programs with new variables and values More time is needed for review because of the complexity (real change due to change in question vs. change due to processing or other error)

4. Data Products Dissemination

In this step, staff deliver the final data to the Center for Enterprise Dissemination Services and Consumer Innovation, stage the data for release, and publish data tables to data.census.gov. Additionally, updates are made to the ACS website with release notes, updated sample forms with the new content, and other documentation.

System/Operation	Activities
Data Dissemination System	Update user notes and links on tables Update documentation (sample forms, subject definitions, comparison guidance, summary of table changes, code lists, etc.)

7. COMMUNICATIONS

The ACS program depends heavily on engaging stakeholders in the development of the program and seeking stakeholder input as much as possible in decisions affecting ACS data products. Typically, consultations with stakeholders begins early in the ACS development process to gain feedback on the overall approach and identify potential pitfalls and obstacles. Implementation in 2027 allows the Census Bureau to properly engage with stakeholders to continue building and maintaining trust.

Staff will conduct their usual outreach associated with the 1-year and 5-year data releases but also expect that additional content may be developed for the website and additional presentations may be held to explain the changes in the data products and data itself.

8. IMPLEMENTATION DEPENDENCIES AND CONSTRAINTS

In addition to the activities highlighted above that need to be done to implement SPD 15, other Census Bureau priorities and projects must be considered in determining the timeline for implementation.

1. County-Level Bridging Factors

When OMB publishes new race and ethnicity data standards, federal agencies require data bridging guidance to compare race and ethnicity data collected under the new standards with data collected under the old standards, and vice versa. “Bridging factors” are the proportion of a race and ethnicity category from one classification scheme that can be assigned to a corresponding race and ethnicity category in another classification scheme. Counts or estimates of a race or ethnicity category in one classification can be multiplied by a bridging factor to calculate a comparable count or estimate of that population in another classification. The development of race and ethnicity data bridging methodology, enabling the comparison of data collected under 2024 SPD 15 with data collected under 1997 SPD 15, is critical for federal agencies and the public.

An Interagency Technical Working Group (Working Group) of Federal Government career staff who represent programs that collect or use race and ethnicity data was convened by the Chief Statistician of the United States.³ The Working Group’s Bridging Team developed national-level bridging factors as part of the release of the updated SPD 15. Future iterations of the Working Group’s bridging factors and programs are expected to be more expansive and flexible, allowing data users to bridge race and ethnicity distributions separately by state and county, and to incorporate other stratifying variables. This important work will help to the inform the ACS implementation of SPD 15.

2. Updated Population Estimates

The ACS has historically used ratio estimation to take advantage of independent population estimates by sex, age, race, and Hispanic origin, and estimates of total HUs produced by the Population Estimates Program of the Census Bureau. This results in an increase in the precision of the estimates and corrects for under- or over-coverage by geography and demographic detail. This method also produces ACS estimates consistent with the population estimates by these characteristics and the estimates of total HUs for each county in the United States. (For more information about how population estimates are used see Chapter 11 of the ACS and PRCS Design and Methodology report.)

Updated population estimates that align with SPD 15 and the combined race/ethnicity data framework, along with county-level bridging factors, are needed as inputs into the ACS weighting and estimation process to apply population controls.

3. Race/Ethnicity Coding Improvement Project

The Census Bureau is currently undertaking its Race/Ethnicity Coding Improvement Project, which aims to enhance and improve the code list that was used in the 2020 Census ​to ensure that detailed write-in responses to the race and ethnicity question are accurately coded and tabulated​. The ACS uses the same code list as the decennial census for coding and tabulating data on race and ethnicity.

A critical component of this project is public feedback and stakeholder engagement. In Fall 2024, the Census Bureau plans to publish an FRN seeking public feedback on the proposed race and ethnicity code list. It is anticipated that there will be a large volume of feedback, which subject matter experts will need to review, and adjudicate before final decisions on the code list are made. Once final decisions are made, updates will need to be made to the code list and Masterfile used for the coding operation and to data product specifications.

4. Survey and Enterprise Innovations and Initiatives

The Census Bureau is transforming and modernizing our approach to creating statistical data products. Four key innovation areas have been identified to allow us to respond more effectively to the rapidly evolving needs of the nation. These innovation areas include transitioning to a modern platform for data collection, using administrative data to replace survey questions and to develop new data products, modernizing the platform for data dissemination, and using cloud-native software and open-source technologies to process and analyze data.

These critical innovations and initiatives require staff to switch all ACS data collection instruments and associated functions to the Census Bureau’s new data collection system, transition ACS systems to open-source software (i.e., refactor millions of lines of code), and move from server-based platforms for data access and transfers to a cloud-based system, all while continuing to maintain the current survey and innovate on other aspects of the survey including administrative data use and content testing. These transitions are scheduled to occur in the next few years and must be completed well before the 2030 Census production operations. 

5. 2025 Content Changes

The approved content for the 2025 ACS and PRCS reflects the changes to content and instructions that were recommended because of the 2022 ACS Content Test. The scope of a content test is to determine question wording, instructions, response categories, and the underlying constructs to improve the quality of data collected. The 2022 ACS Content Test results led to recommendations to implement changes for the household roster, educational attainment, health insurance coverage, and labor force questions. Additionally, three new questions are being added: sewer, electric vehicle, and solar panels.

Overall, these 2025 content changes impact the ACS data collection instruments but do not require extensive post-processing or data product changes like the change to the race and ethnicity question:

• Household Roster: Requires changing instructions on the paper questionnaire but does not change any downstream activities (variables are not changed); electronic instruments require questions and pathing changes, but summary outcome variables are the same and do not require downstream changes. There are no changes to data products because of this question change.

• Educational Attainment: The revision changes formatting and wording to clarify the response options. This requires updating instruments and collapsing categories in post-data collection processing. There is no coding for this question. Categories will be collapsed for data products, which are relatively straightforward to define.

• Health Insurance Coverage: Changes to the health insurance coverage question include a change in formatting of the question that adds an explicit response category for those who are uninsured, reordering some response options and rewording response options for some categories. This requires updating instruments and post-data collection variable updates that are easily mapped to existing categories. There are no changes to coding procedures. We are still determining what changes are needed for edits and data products.

• Labor Force: Labor force questions related to when the person last worked, the number of weeks, and the number of hours worked are being updated to clarify instructions to only include work for pay, to include all jobs a person may hold, and to ensure that military service is included. There are no changes to variables, post-processing steps, or data products.

• Sewer, Electric Vehicles, and Solar Panels: Requires implementation of new questions and post-data collection processing that were tested and used in the 2022 ACS Content Test. These questions have no coding. New edit and data products will be created, but because there are a limited number of categories for each question (2-3), the data products are relatively straightforward to define and are small in number.

6. Next ACS Content Test

Planning for the next ACS Content Test was scheduled to begin in the spring of 2024. However, based on lessons learned from the 2022 Content Test revisions to the Content Test process need to be considered before next steps are taken.

The process for altering ACS content is designed to help ensure that any proposed changes are necessary and will produce quality, useful information for the nation. Adding a question or making a change to the ACS involves extensive testing, review, and evaluation over a 5-year period. Over 25 federal agencies participate in this process and make recommendations to the Interagency Council on Statistical Policy’s Subcommittee on the ACS (ICSP-SACS). An important step in the process is providing an opportunity for the public to provide feedback. We do this through publishing a series of FRNs that invite public comments on proposed changes and methods.

Based on the feedback from the FRN from our last content test related to the recommendation to change the ACS questions on disability, the Census Bureau will recommend the ICSP-SACS consider revisions to the protocols to enhance collaboration with data users and other stakeholders at the beginning of, and during, the ACS content change process. Continued work on this process change has been paused due to the focus on meeting requirements for SPD 15. The next steps in ACS Content Test will be delayed until the ICSP-SACS has an opportunity to review and provide feedback on the proposed changes and the Census Bureau and OMB agree on the modifications.

7. Disability Stakeholder Engagement Activities

From the FRN on the proposed changes to the 2025 ACS content, the majority of the 12,000 comments expressed concerns with the proposed changes to the disability questions. Based on this feedback, the Census Bureau has retained the current ACS disability questions for the 2025 ACS. The Census Bureau has been leading a stakeholder engagement effort with the disability community to address their concerns. The Census Bureau, National Center for Health Statistics (NCHS), OMB, the Leadership Conference on Civil and Human Rights, the National Council on Disability, the Consortium of Constituents with Disabilities, and the National Disability Rights Network held a meeting on September 30, 2024, with federal agency disability stakeholders, disability community representatives, data users, researchers, and disability advocates. The purpose of the meeting was to discuss data needs and data uses surrounding the topic of disability.⁴

The OMB has established the Disability Data Interagency Working Group (DDIWG) to inform the Federal Evidence Agenda to advance disability equity. The DDIWG’s work will inform federal surveys and administrative collections on how to collect data on disability. The DDIWG is tasked with identifying the needs for disability data in the Federal Government, identifying resources that are currently available in the Federal Government on disability data, and identifying gaps in the data. The DDIWG work will set up priorities for federal agencies. The DDIWG is exploring what is in scope for the Evidence Agenda, it is reviewing inputs from community engagement, the DDIWG will explore whether it will define disability and will it define how to measure disability. The DDIWG will provide recommendations to federal surveys to advance disability equity. A report is expected by fall of 2024.

A potential Committee on National Statistics (CNSTAT) panel would build on the Federal Evidence Agenda produced by the Disability Data Interagency Working Group (DDIWG) to advance disability equity. Using its expert knowledge, a CNSTAT panel would be useful to the Census Bureau and the Federal Government in continuing the data work outlined by the DDIWG. Given the data needs and gaps identified by the DDIWG, the panel would provide a recommendation for the best approach for collecting disability data, whether that’s how to collect it on existing surveys or proposing new approaches for collecting the data. The panel would make recommendations for studies to be conducted and future work to continue development of disability statistics.

8. Sexual Orientation and Gender Identity Content

The Census Bureau is seeking approval to conduct the 2024 ACS Sexual Orientation and Gender Identity (SOGI) Test. The Census Bureau proposes testing questions about sexual orientation and gender identity to meet the needs of other federal agencies that have expressed interest in or have identified legal uses for the information, such as enforcing civil rights and equal employment measures. The information collected in the 2024 ACS SOGI test will be used to evaluate the quality of data from questions on sexual orientation and gender identity and to inform recommendations about whether and what questions to implement in production ACS.

Presentations about this test and the next steps have indicated the earliest possible year SOGI questions could be implemented on the ACS is 2027. This timing accounts for when test results will be available and when we could get OMB clearance to add the questions to the survey. Additional assessments are still needed about the impacts to coding, editing, and data products, especially for sex data which are used as cross-tabulations and as edit inputs for other topics.

There are no final determinations of the SOGI implementation year. The SOGI implementation timeline would have to consider other competing survey changes and program priorities, such as SPD 15 and enterprise innovations (discussed in Section 8.4).

We anticipate that adding new questions on sexual orientation and gender identity will create new edits and tables. While sex is a foundational item like race and ethnicity, with fewer categories there are fewer and simpler impacts on edits and data products than the race changes. It is not clear yet what, if any, impact there will be on the edits.

Additionally, these new question changes are limited to the ACS, while SPD 15 affects all Census Bureau surveys that collect race and ethnicity data. Therefore, subject matter experts on race and ethnicity have a more significant demand on resources to update specifications and be involved in testing systems for a multitude of surveys.

9. CONCLUSIONS

Nearly every step in the ACS survey life cycle, from content development through data product identification and testing, requires staff to update requirements, specifications, program changes, and test the changes. The changes are complex, dependent on other projects and activities, and in some cases need to waterfall from each other. The associated risks to data quality and heightened errors are of sufficiently grave concern that a deliberate, thoughtful approach must be followed.

While prior content testing provides a solid basis for the questions and associated variables for data collection instruments, additional time and resources are needed to develop the editing, imputation, and data products for the new combined race/ethnicity question.

Based on our assessment, the Census Bureau plans to implement the updated race and ethnicity standards into the 2027 American Community Survey data collection cycle.

10. REFERENCES

Groves, R. M., F. J. Fowler, M. P. Couper, J. M. Lepkowski, E. Singer, and R. Tourangeau, (2004). Survey Methodology. Hoboken, NJ: John Wiley & Sons, Inc.

Office of Management and Budget (2024). “Updated Statistical Policy Directive No. 15.” Retrieved June 27, 2024 from https://www.federalregister.gov/documents/2024/03/29/2024-06469/revisions-to-ombs-statistical-policy-directive-no-15-standards-for-maintaining-collecting-and

U.S. Census Bureau (2022). “American Community Survey and Puerto Rico Community Survey Design and Methodology (v3.0).” Retrieved June 27, 2024, from https://www.census.gov/programs-surveys/acs/methodology/design-and-methodology.html

1. Preliminary SPD 15 Implementation Schedule for ACS

ACS Process	ACS Operation	Activities	Timeline for 2027 Implementation
n/a	OMB Clearance	Question and Help Text Wording	Summer and Fall 2025
Data Collection and Capture	ACS Case Management and Control System	Requirements Specifications, System changes, Testing, and Production Release	Spring 2026 Summer – Fall 2026
Data Collection and Capture	Language Support	Translation for all modes (question, help text, guides) Training materials and manuals for all modes	Winter 2025 – Spring 2026 Fall 2026
Data Collection and Capture	HU Mailing	Paper questionnaire changes & printing	Fall 2025 – Summer 2026
Data Collection and Capture	HU Internet Data Collection	Instrument changes & testing	Fall 2025 – Summer 2026
Data Collection and Capture	Data Capture	Requirements Paper data capture changes	Spring 2026 Summer 2026
Data Collection and Capture	Failed-edit Follow-up	Instrument changes & testing Training and manuals	Fall 2025 – Summer 2026 Summer – Winter 2026
Data Collection and Capture	HU Personal Interview	Instrument changes & testing Training and manuals	Fall 2025 – Summer 2026 Summer – Winter 2026
Data Collection and Capture	Telephone Questionnaire Assistance	Instrument changes & testing Training and manuals	Fall 2025 – Summer 2026 Summer – Winter 2026
Data Collection and Capture	GQ Paper and CAPI Data Collection	Questionnaire and instrument changes & testing Training and manuals	Fall 2025 – Summer 2026 Summer – Winter 2026
Data Collection and Capture	GQ Internet Data Collection	Instrument changes & testing	Fall 2025 – Summer 2026
Post Data Collection Processing	Data Integration	Requirements, Specifications, System changes, and Testing	Summer 2026 – Spring 2027
Post Data Collection Processing	Write-In Response Coding	Development and revision of the code list and Masterfile updates, including public comment	Spring 2024 – Summer 2026
Post Data Collection Processing	Write-In Response Coding (automated coding and clerical coding)	Requirements, Specifications, System changes, and Testing Training and Materials Development & Coder Training	Fall 2026 – Spring 2027
Post Data Collection Processing	Edits	Research and develop new race and ethnicity edits	Spring 2024 – Fall 2027
Post Data Collection Processing	Edit Input	Specifications, system changes & testing	Fall 2027 – Spring 2028
Post Data Collection Processing	Editing, Allocation, and Review for 1-year Data	Specifications, system changes, testing, and production deployment	Summer 2027 – Summer 2028
Post Data Collection Processing	Disclosure Avoidance	Specifications, System changes, Testing, and Production deployment	Fall 2027 – Spring 2028
Post Data Collection Processing	Weighting and Estimation	Specifications, system changes, Testing, and production deployment	Fall 2027 – Spring 2028
Data Products	Data Reasonableness Review	Specifications, system changes, Testing, and production deployment	1-year: Winter 2027 – Summer 2027 5-year: Winter 2027 – Fall 2027
Data Products	Data Product Development	Research and develop new data products	Winter 2026 – Winter 2027
Data Products	Data Products Development [Standard Data Products & PUMS]	Specifications, System changes, and Testing	Summer 2027 – Spring 2028
Data Products	Data Products Review [Standard Data Products]	Production data products review	1-year: Spring 2028 – Summer 2028 5-year: Summer 2028 – Winter 2028
Data Products	Data Products Dissemination [Standard Data Products & PUMS]	Dissemination of products and update of website documentation	1-year: Spring 2028 –Fall 2028 5-year: Fall 2028 – Winter 2029

1 For ease of discussion, the term ACS is used throughout this document to represent both the ACS and PRCS.

2 Note that references are made throughout this document to the ACS, but all changes also apply to the Puerto Rico Community Survey (PRCS).

3 U.S. Office of Management and Budget Interagency Technical Working Group on Race and Ethnicity Standards. (2023). spd15revision.gov. Retrieved from: https://spd15revision.gov/content/spd15revision/en/about.html.

4 See Livestream Meeting With Stakeholders on Disability Data Needs (census.gov) for more information about this meeting.

6