danger
Election Day is tomorrow, November 5

Ballots must be postmarked by Nov. 5, 2024, or dropped off at a ballot box by 8 p.m. tomorrow. Find your nearest ballot box

Race, Ethnicity, Language, Disability, and Tribal Affiliation Demographic Data Standards Guidance

Information
This guide helps City of Portland staff implement the Race, Ethnicity, Language, Disability, and Tribal Affiliation Demographic Data Standards (RELDTA) in Civil Rights Administrative Rule ADM-18.03
On this page

Guide Overview

This guide helps City of Portland staff implement the Race, Ethnicity, Language, Disability, and Tribal Affiliation Demographic Data Standards (RELDTA) in Civil Rights Administrative Rule ADM-18.03

Use of these data standards provides more accurate and representative descriptions of the populations served by the City of Portland. This is key to understanding disparities, Civil Rights Compliance, and providing equitable access to City resources and services. Historically, City staff have not aligned to a single standard. This limits the ability to compare data, can lead to data quality issues, and increases workload and burden on City staff and partners collecting and analyzing data. For more details on the purpose of demographic data and standards, refer to the Civil Rights Administrative Rule ADM-18.03.

The City of Portland RELDTA standards incorporate local, state, and national best practices. Lessons learned from City implementation, community partners, regional, and federal guidance will continually be incorporated into this guide. The published date appears below the title. As new versions are published, changes to the guide are documented in appendices. 

City staff should use the guide to learn about effective use of the standards. Collaborating with communities and staff with equity, data analysis and data management subject matter expertise throughout the data life cycle is key to effective implementation. 

For additional support to implement the standards or share lessons learned to improve this guidance, contact: OEHR@portlandoregon.gov.


Using the Standards

When to use and not use the standards?

Civil Rights Administrative Rule ADM-18.03 applies to all City staff collecting demographic information about the public. The rule also applies to all contractors, subcontractors, and grantees working on behalf of the City collecting demographic information about the public. Admin Rule ADM-18.03 describes examples of the public as including, but not limited to, people accessing City services or programs, people participating in community engagement with the City, and people working in partnership with the City through contracts or grants.

Civil Rights Administrative Rule ADM-18.03 includes an exception if there are alternative federal, state, or other required data collection and reporting standards already in place. When City staff have preempting rules, the staff member responsible for the data must also maintain documentation of the alternative standards. Bestefforts should also be applied to map the alternative standard back to the City standards. For example, the comprehensive standards aggregate or rollup to the minimum standards. See the ”How do I rollup, map, or aggregate different data standards” section of this guide for more information. 

For all data collections, City staff must assess if it is appropriate, feasible, and safe, to ask about demographic data. We need to balance several considerations. These include balancing the need to understand the populations we work with and serve to identify and address disparities against the appropriateness of collecting data from specific populations based upon several factors ranging from historical relationship with the City to collection sample size. Staff must consider what questions they are looking to answer through data collection along with the City's overall purpose for collecting demographic data. 

When consulting with analysts, researchers, community members, tribal nations, or other data subject matter experts in the collection of data, city staff must proactively work to maintain confidentiality and privacy. These same partners may help staff assess if a data collection may be considered statistically unreliable or significantly linked with abuse or misuse of data. If collecting one or any combination of demographic categories compromises privacy, that data should not be collected or reported without appropriate context, cautions, and safeguards in place.

There may be scenarios when a survey, project or program focuses on a specific population. In this scenario, that demographic variable would be excluded but other standards would still apply. For example, a survey may specifically focus on those who primarily speak Spanish at home and Spanish is their preferred language. Language data would not be collected since it is not needed. Race, ethnicity, disability, and tribal affiliation would. City staff may partner with a tribal government for a data collection. American Indian or Alaska Native tribal affiliation may not be necessary. 

Identities are fluid and complex. Before excluding variables, staff should consider and weigh benefits, check sovereignty agreements, and consult with partners, data support staff, and/or Office of Equity staff. Special exceptions may apply. 

When should I use the minimum or the expanded standards? 

The standards include both a minimum and expanded version for race, ethnicity, disability, and language. Staff must weigh privacy, needs, use, sample size, and context to determine whether to use the minimum, expanded, or a combination of standards. 

Where possible and necessary, consider looking at the most complete and detailed picture of communities at the subcategory level. Disparities exist within racial and ethnic categories. It is important to include, understand, and be transparent of the distribution of responses and of subpopulations. Does your project or program have a specific data collection need to understand the expanded race and ethnicity of those that you serve? Does your project also have a large sample size so the risk to identify people is lower? Then the expanded race and ethnicity standard may be the best fit. However, if you identify privacy, confidentiality, and anonymity concerns, the minimum standards may be the best fit.

In some scenarios you may collect the expanded data to fit your data collection needs but choose to report out at the condensed level. See the “How do I rollup, map, or aggregate different data standards?” section for more information. 

Can you reformat or move questions around? 

Generally, yes. City staff who collect these data can adjust the order of the questions/sections to fit the communities they serve.

Can we rephrase a question and/or categories?

Generally, no. While you may want to rephrase a question to make it easier to read or rephrase categories to be responsive to specific feedback, this can change how respondents answer. This makes it difficult to compare and combine datasets collected by others in the City. That being said, the standards do aim to be inclusive and will be reviewed periodically, understanding that identities are fluid and personal.

As an example, we use the term "Latino/a/e/x or Hispanic" as an inclusive alternative to "Latino" alone. Some individuals may prefer one part of the term versus another. The standard includes multiple terms and is the language that should be used when collecting this information. In Spanish, the masculine version of a noun is gender neutral, but some advocates, mostly based in the United States, have called into question the neutrality of masculine nouns. Language evolves and more inclusive alternatives to “Latino” have included Latina/o and Latin@. We use Latino/a/e/x to intentionally include gender nonconforming and nonbinary people of Hispanic or Latin American origin who might not identify with the masculine or the feminine. Latinx is a relatively new term believed to have originated in the Latino LGBTQIA+ community. While not without dissent, Latinx is being increasingly adopted by advocates, universities, and even government agencies. 

Community and staff feedback from implementation of the standards will be used to identify future changes needed. If you have identified a specific need to rephrase or adjust a question or category, consult with Office of Equity to identify a solution.


Collecting and Analyzing Demographic Data

What to consider when designing RELDTA data collection?

Race, ethnicity, language, disability, and tribal data must be collected, analyzed, and reported in an accurate and useful manner. This begins with designing and planning the data collection methods, settling how the data will be stored and managed, determining analysis and reporting needs, and considering any community concerns such as privacy, potential misuse and community ideas about data uses. 

The standards have both a minimum and expanded set of standards under each category except tribal affiliation. City employees should weigh privacy, needs, use, size, context, etc., to determine whether to use the minimum, expanded, or combination of standards. The following sections include additional considerations.

Planning how data will be used

  • Review any additional reporting requirements from funders, agencies, community partners, etc. 
  • Collaborate with communities to understand privacy concerns, questions about use or misuse of data, or other helpful data and context to include.
  • The minimum and expanded standards represent baseline standards. You are allowed and encouraged to collect additional data as needed.
  • Optional questions are included depending on desired information, need, subjects, and style. When the question is left open ended rather than predefined, be prepared to code to condensed or aggregated standards for analysis and reporting.

Data management

  • Collaborate with data analysts, data governance staff, and/or City Attorneys to ensure the categories can be collected while maintaining privacy and information security priorities and other data management requirements. 
  • If you’re working with a contractor to collect data on your behalf, make sure they understand the standards and you have clear data sharing agreements to manage the data. 

Formatting

  • Ask demographic questions at the end of your survey.
  • Format questions so the respondent can easily see the question and all response choices at once (e.g. on the same page of a paper survey). This helps the respondent feel comfortable, while also helping to reduce misclassification by having all options available in a single location.
  • Tribal affiliation is not the same as race and ethnicity and should be presented as a separate question. However, topics and identities overlap. We recommend presenting the tribal affiliation question if “American Indian, Alaska Native, or Indigenous Peoples of the Americas” is selected in the race/ethnicity question. If using a tool without the capability for survey logic or to nest questions, the tribal affiliation question should be included as a separate question.  
  • Requests for demographic information must be distinct/separate from questions related to program eligibility criteria. An individual’s decision not to answer questions related to demographic data shall not affect eligibility for public programs. 
    • If you have a program with specific eligibility requirements related to demographics, work with data analysts, data governance leads, City Attorneys, and other subject matter experts to design methods for verification that lead with trust and support for self-attestation.

How do I protect the respondents’ privacy?

All records prepared, owned, used, or retained by the City are considered public records under state law and may be subject to disclosure. City employees should make every effort to collect demographic information that cannot be used to identify the respondents in order to protect their privacy. Do not attach a name, address, phone number, or email address to a demographic form. Consult with staff with data analysis and data management expertise to develop data collection methods so privacy is built in from the beginning.

Every effort should be made to make all demographic information anonymous. If it cannot be anonymized, connect with analysts, managers, data governance leads, and attorneys before collecting it to discuss your project’s needs and strategies for confidentiality, protection, and voluntary disclosure.  

Even if you do not collect names or addresses in your survey questions, it still may be possible to identify the respondent if they provide their gender, race, age, zip code, housing type, etc. Take, for example, the only apartment complex in a particular ZIP code. Demographic data collection would identify a respondent as living in that ZIP code and in an apartment. If that apartment only has 10 units, one could determine with reasonable accuracy where that person lives. If you also asked about commute information you might know when they are regularly home or not. This presents issues around protecting individual privacy that must be considered when designing data collection. The more demographic questions the City asks, the higher the privacy risk.

All City employees must adhere to the requirements of the Privacy Principles and BTS-2.01 Information Security Admin Rules to guide their actions when collecting and using the public’s personal information. These measures for data security and respondent privacy defend against growing threats in cybersecurity and data misuse. It is also important to communicate the efforts and gravity of these measures to build public confidence and trust in the City’s data management to ensure high response rates.

What to consider when collecting RELDTA data?

When you are conducting a survey and collecting demographic information, the following principles apply: 

  1. Requests for demographic information shall be voluntary, and an individual has a right to decline to answer any or all demographic questions. Do not override an individual’s right to refuse to report for any or all the questions. All declined to answer responses/submissions or missing demographic information will be identified and recorded in the data collection for analysis. 
  2. Self-identification is the standard. A key principle underlying RELDTA is that of self-reporting. The City and City contractors, subcontractors, and grantees should ask individuals to self-report and should not assume or judge ethnic and racial identity, preferred signed and spoken language, disability status, or American Indian or Alaska Native tribal affiliation, except as mandated by state or federal reporting requirements. Since the RELDTA data standards reflect identities and preferences, self-reporting will also provide the most accurate information.
  3. Individuals have the right, and shall be offered the option, of selecting or providing one or more (multiple) racial or ethnic, language, disability, and American Indian or Alaska Native tribal affiliation designations or responses. 
  4. Identities are fluid and personal. Prioritize respondents’ comfort. Providing an option to write in an identity not listed helps relieve pressure and helps the City track emergent populations.

How do I tally and present information when a respondent selects multiple responses? 

In self-identification of demographic information, respondents can choose to check off the boxes of more than one response. There are a variety of ways to analyze data where there are multiple responses from single respondents. 

Using race/ethnicity responses as an example, here’s an illustration of how to analyze data where some respondents provide more than one response for a single question. This example can be extrapolated to apply to other demographic categories as well. 

When tallying the counts of each race/ethnicity, often figures for people who mark/choose only one racial identifier are reported under “alone” statistics. People who choose more than one identifier are reported under “alone or in combination” stats or figures. The Oregon Health Authority REALD Implementation Guide provides an overview of different types of analysis and reporting and associated risks. 

We also want to respect and consider recommendations from community partners by using alone and in combination where possible (and while weighing risks). The Coalition of Communities of Color notes in their Leading with Race: Research Justice in Washington County report that “We use ‘alone or in combination with other races’ rule to collect data about all communities of color. This means that biracial and multiracial people are counted as belonging to each community that they identify with. These biracial and multiracial people disappear when researchers and policy makers use our ‘alone’ figures to define the size of our communities, and their experiences get obscured, rendered invisible, and denied.”

When creating summary values or figures about race and ethnicity data, include this context about your methods. It is okay to note and be clear that total counts in a bar plot or bar chart of race and ethnicity data may include duplicated individuals served because individuals can choose multiple categories. Identify and share other variables collected that do report unduplicated individuals served. 

Formatting data for multiple response questions will differ based on your survey tool or data collection methods. To prepare for these analyses, it may be helpful to run a test survey to understand the format of responses. This can help you prepare any data cleaning scripts or processes needed before analyzing the data. In one example of implementing these standards by the Rescue Plan Data and Equity Strategies Team, the data were primarily received in a table/csv/excel format. The team provided the following formatting requirements to help facilitate consistent submissions across more than 40 projects and to make data cleaning prep for analysis more efficient: 

When submitting data in a table/csv/excel format, use a comma to separate multiple selections of categories by one person. 

For example, this person has self-reported three race and ethnicity standard categories to describe their racial or ethnic identity: 

race_ethnicitydescription_race_ethnicity
Asian, Black or African American, White 

When an individual chooses “Not listed above, please describe”, record that text as the category selection. Create a separate column to record the open-ended text description. An example name of this new column is “description_race_ethnicity”. 

For example:

race_ethnicitydescription_race_ethnicity
Not listed above, please describeopen-ended text description provided by person

How do I analyze tribal affiliation responses? 

As we begin to collect tribal affiliation data, we are learning more about how to analyze and summarize these data. The first step is to familiarize yourself with the American Indian or Alaska Native Tribal Affiliation definition in Admin Rule ADM-18.03. This definition was built with input from the City’s Tribal Relations Program. 

One common issue can be misunderstanding of the question by the respondent. For example, someone may enter a response such as “I’m married to a Native American” for the category/open ended question of “Enrolled Member. Tribal Affiliation(s):”. This does not meet the definition of Enrolled Member and does not describe the name of the American Indian Tribe or Alaska Native Village Corporation. Another example is a respondent may enter a response such as “My great, great, great grandmother” for the category/open ended question of “Descendant. Tribal Affiliation(s):”. This does not meet the definition of a descendant in the first or second degree (parent or grandparent) and does not describe the name of the American Indian Tribe or Alaska Native Village Corporation. In both examples, the responses would need to be correctly put into the category response of “I am not an enrolled member or descendant of a federally or state-recognized American Indian Tribe or Alaska Native Village Corporation.”

Work with the Tribal Relations Program or other subject matter experts in your teams with knowledge of tribe names to further understand valid answers. This data review is necessary before being able to summarize response counts or create figures to report out and share back about tribal affiliation of your respondents. 

How do I rollup, map, or aggregate different data standards? 

The expanded and minimum categories align with national and local community best practice, the State of Oregon’s ORS 943-070 REALD standards, and can be aggregated or rolled up to the US Office of Management and Budget Standards (established by Directive No. 15). These are examples of how to map the City of Portland standards to other established data standards. See the Aggregation Crosswalk for these relationships. 

Office of Management and Budget (OMB) establishes federal standards for race and ethnicity. Data reported to other agencies, such as the federal government, may need to be rolled-up into the OMB standards for race and ethnicity.

There are multiple rollup schemes for aggregating granular data with OMB standards. Please look at methodology and coding reports from the Census Bureau, ACS PUMS, or consider using the CDC/HL7 code set. There are over 900 race and ethnicity codes in the CDC/HL7 Code Set introduced in 2000.


Statements to include when collecting demographic data

Introductory statement

An introductory statement or message to respondents helps improve expectations and understanding about the data being requested, how the data will be used, and potential disclosure of their responses. Include the following information:

  • Purpose of collecting demographics. Examples:
    • “The City of Portland is committed to diversity, equity, justice, and inclusion and will use any demographic information you provide to help ensure equitable representation in the work we do, to provide the best services and policies, and to reduce inequities and disparities.”
    • “We ask about race, ethnicity, language, disability, tribal affiliation, and gender [add in other categories asked] to ensure equitable representation in the work we do, to provide the best services and policies, and to reduce inequities and disparities.”
  • Nondiscrimination Statement (if the demographics are part of any application, provision of services, or decision-making process).
    • “State and federal law prohibit the use of this information to discriminate against you.”
  • Transparency
    • Provide a concise statement about how your project or team is planning to use the data and/or provide a report on the data after it is collected.
    • If possible, provide a link where they can find out more information.
  • Voluntary
    • Make sure to indicate that completion of demographics is completely voluntary and that the respondent has the right to decline to answer any questions. This can include letting respondents know there is an option of “I prefer not to answer” for all demographics-related questions.
    • Example of purpose, nondiscrimination and voluntary statements combined:
      • “Completion of this section is completely voluntary. The City of Portland is committed to diversity, equity, justice, and inclusion and will use any demographic information you provide to help ensure that the City’s services reach a broad cross-section of the community. State and federal law prohibit the use of this information to discriminate against you.”
  • Disclosure
    • Before including disclosure clauses with indications that information can be kept confidential, check with the City Attorney’s Office and have a documented plan for how you will manage these data and any requests. This is especially important for disclosure clauses that give individuals the option to check yes or no about how they want their information disclosed or if respondents are given an option to waive confidentiality. You need to make sure you legally and technically can deliver those options.
    • Here is an example of a disclosure statement to provide transparency about public records law:
      • “Information you provide to the City of Portland on this form may be subject to disclosure under state law such as Oregon’s public records law.”

Meaningful Access Statement

Any data collection efforts involving the public should include a Meaningful Access Statement. This should be included for the entire survey, not necessarily in an introduction to the demographics section. 

Example Civil Rights and Meaningful Access Statement: The City of Portland is committed to equity and meaningful access, and prohibits discrimination on the basis of race, color, national origin, Limited English Proficiency, disability, age, sex, religion, income level, sexual orientation, gender identity, familial status or other protected class as provided by Title VI of the Civil Rights Act, Title II of the Americans with Disabilities Act, and related authorities. To request translation, interpretation, accommodations, modifications, or other auxiliary aids or services, or to file a complaint of discrimination, contact 503-823-4000 (311), Relay Service & TTY: 711, or 503-XXX-XXXX.