ID Verification

Overview

Unemployed workers should have straightforward access to unemployment insurance (UI) benefits while they are looking for new jobs. However, in recent years workers face a new obstacle: Burdensome identity (ID) verification procedures that make it difficult, if not impossible, for many workers to access the benefits they are entitled to.

ID verification measures can be a particularly high barrier for Black, indigenous, and other workers of color, immigrant workers, disabled workers, and underpaid workers because systems are often set up based on assumptions about technology access, credit history, language, digital literacy, and appearance that are more likely to apply to well-paid, highly educated, non-disabled, English proficient, and white workers.

ID verification is intended to reduce identity fraud by ensuring that a person attempting to claim UI benefits is who they say they are. In December 2020, Congress enacted an ID verification requirement for pandemic benefits after criminal enterprises used stolen identity data to fraudulently claim pandemic UI benefits. Even though identity fraud for regular UI benefits was rare prior to the pandemic, most states instituted ID verification systems for regular benefits as well as the pandemic benefits which have now expired.

Overwhelmed by UI claims during the pandemic, under-resourced state UI agencies often adopted new ID verification systems hurriedly, turning to relatively untested systems without considering the challenges they might pose for UI access and equity. For example, a little-known company called ID.me got its first contract to provide ID verification services to Florida’s state UI agency in June 2020. A year later, the company had contracts with at least 25 states, despite numerous concerns about due process, privacy, and discriminatory impacts.[4]

State ID verification procedures

States have a range of choices about ID verification including whether to require an ID verification process at all, which workers must undergo the process, when in the application process to require it, what methods workers can use for identity proofing (including options to prove their identity off-line, for example by mailing documents or presenting them at an in-person location), and whether and how to use private vendors.

In October and November 2022, researchers with the Digital Benefits Network at the Beeck Center for Social Impact + Innovation at Georgetown University compiled data on ID verification steps for online unemployment insurance applications based on publicly available information.[5] They documented the following state practices:

• Requiring ID verification:A total of 27 state agencies required workers applying for UI benefits to actively prove their identities at some point before, during, or following the claims filing process; six agencies did not appear to require workers to prove their identities; and five agencies required identity proofing only in some cases. For 15 agencies, it was unclear whether a program required identity proofing, or whether they required it for all workers applying for UI.
• Implementing ID verification at different stages of the UI application process: Of the 27 agencies that required identity proofing, 10 prompted workers applying for UI to complete identity proofing actions before initiating an application and 10 incorporated identity proofing duringthe application process. Nebraska required identity proofing actions after application submission. Iowa allowed workers who decline or fail identity proofing during the application process to submit identity documentation online, by mail, or in person. For five agencies that seemed to require active identity proofing actions, it was unclear when identity proofing would take place.

• Using various methods of ID verification: The most common ID proofing method, used by 22 agencies, involved uploading identity documentation and a video selfie, with the state using facial recognition technology. In addition, 10 state agencies used knowledge-based verification as an identity proofing method; nine agencies allowed workers applying for UI to mail, fax, or present identity documentation in person; and five state agencies offered workers choices in identity proofing methods. Researchers note that they did not track what happened when the ID proofing process failed: Additional states may provide non-digital options in these cases.

• Contracting with vendors providing ID verification services: Researchers at the Beeck Center documented 27 state agencies that contracted with the private identity verification provider ID.me at some point, although several agencies appeared to have discontinued the service. At least four state agencies (Iowa, Kansas, Rhode Island, and Wisconsin) contracted with LexisNexis Risk Solutions for knowledge-based verification services. Other agencies used state-specific identity verification tools such as Illinois’s Department of Economic Security, which used ILogin (the state’s service) for sign-in and identity proofing. (ILogin uses the private company Experian for identity proofing.) The Oklahoma Economic Security Commission used a customized identity proofing solution, VerifyOK, built by IDEMIA. As described further below, Arkansas and Hawaii are now using the federal government’s Login.gov and U.S. Postal Service for ID verification.

For additional details, including which states followed each procedure, see the Beeck Center’s resource Digital Authentication and Identity Proofing in Unemployment Insurance. Because the resource is updated periodically, numbers may change.

U.S. Department of Labor offers support for Login.gov and the U.S. Postal Service

As a public alternative to unaccountable private vendors, the U.S. Department of Labor (USDOL) announced a National ID Verification Offering in July 2023.[6] USDOL is enabling state agencies to use the government-operated system Login.gov developed by the U.S. General Services Administration for online ID verification. USDOL also provides states with an in-person option: In states that choose to participate, workers who have difficulty using a technology-based ID verification system or prefer not to can verify their identities in-person at a U.S. Postal Service (USPS) retail location.[7]

USDOL will use federal funds from the American Rescue Plan Act to cover at least two years of transaction costs for ID verification for participating states. States had until August 18, 2023, to express interest. The program was launched with a pilot in Arkansas[8] and was rapidly expanded to Hawaii in the aftermath of the Maui wildfires.

National Institute of Standards and Technology sets digital identity guidelines

The National Institute of Standards and Technology (NIST) is a federal agency focused on measurement, research, and standards development. The NIST Digital Identity Program sets guidelines detailing a risk-based process and technical requirements for managing digital identities, including identity verification. While NIST does not have regulatory or policymaking authority, policymakers may decide to mandate or recommend the guidelines and standards it issues. USDOL encourages states to comply with NIST standards.[9] NIST is currently in the process of updating its digital identity guidelines for identity proofing, and in December 2022, the agency issued an initial draft for public comment.[10]

Concerns about ID verification procedures

Many of the ID verification procedures currently used by states raise concerns around due process, privacy, and racial equity. These concerns are discussed in greater detail in the section on Research Findings and Arguments to Support Reform. Concerns about specific forms of ID verification, such as facial recognition and knowledge-based verification, are also discussed in that section.

One general concern is that there is currently no federal standard for the type of identification documents allowed[11] and in many states there are not enough types of documents that are allowed as secondary or alternative sources for ID verification. In some states expired IDs are not allowed in any circumstance. This is especially problematic for workers who have employment authorization through covered immigration status. It can also be irrational: just because a person’s authorization to drive has expired, for example, shouldn’t mean that their driver’s license is no longer accepted for purposes of identification.

In addition, some states have overly restrictive ID determinations by agencies and vendors. For example, one worker’s birth certificate was rejected because it had “too many folds to be real.” Another birth certificate was rejected as “hand-written” when all the birth certificates in that state were hand-written at the time of the worker’s birth.

Another concern is that alternative verification methods for workers who cannot or do not want to complete online ID verification are often limited, slower, and less accessible. As a result, workers who tend to face the biggest barriers to online access and have to use the non-digital option—particularly Black, indigenous, and other people of color, limited English proficient workers, immigrant workers, and workers with disabilities—encounter long delays. For example, the private ID verification vendor ID.me recently began to provide an in-person alternative to online verification at UPS stores. However, in practice, once the worker goes in person to the UPS store, they still have to use a computer kiosk to complete the process.[12] While this may be helpful to some workers applying for UI, it does not address privacy or due process concerns around facial recognition.

The recent USDOL ID verification offering concerning Login.gov and the U.S. Postal Service may address some of these concerns in the states that participate, although Login.gov has also raised concerns about biased facial recognition technology and failure to meet NIST identity proofing standards.[13]

Federal Requirements and Guidance

Federal law has long been concerned with verifying the identity of workers seeking UI benefits. Section 1137(a)(1), of the Social Security Act mandates that states require workers to provide their Social Security number as a condition of eligibility for benefits.[14] This means, among other things, that states must have a system to reasonably ensure that the Social Security number provided belongs to the worker applying for benefits. As a result, states have long had some type of system in place to confirm an applicant’s identity. However, the use of widespread digital ID verification is new and federal guidance in this area is still evolving.

Unemployment Insurance Program Letter (UIPL) 16-21 describes identity verification requirements in detail, including acceptable documents to verify ID, the timing of ID verification issues, possible outcomes of an attempt to verify ID, adjudication of ID verification issues and due process standards.[15] The guidance details three times in the UI claims process when ID verification may be used:

• After an application is received by the state, but before the application is entered into the state’s benefit system.
• After a claim is filed, but before payment is issued.
• After a claim is filed and payments have been issued. Because the state either became aware of the ID issue through its normal processes of issue identification or because a financial institution identifies suspicious activity and contacts the state.

The guidance also includes requirements that:

• “A state’s list of acceptable documents to verify ID must include a sufficient variety to provide for equal access to UI benefits for all claimants. For example, if the state requires a state driver’s license as proof for ID verification, it must offer a reasonable alternative for those who may not have that type of identification or provide sufficient time for other means of identification to be obtained.”[16]
• “States must provide alternative mechanisms [to verify identity] for individuals with access barriers, such as a disability or limited English proficiency.”[17] UIPL 22-21 change 2 specifies that “states that utilize ID proofing are required to provide a non-digital alternative option such as in-person ID verification at locations like American Job Centers.”[18]

UIPL 11-23 further details required and recommended fraud prevention and detection activities, including steps to verify workers’ identity and cross-match records.[19]

Identity verification decisions are also subject to the same due process requirements as all other UI eligibility determinations.[20] This includes proper notice and the right to be heard when an identity issue arises. For example, if a state obtains information that raises an issue on whether the name and/or Social Security number used to file a UI claim belongs to the worker who applied for benefits, the state must provide them with proper notice, an opportunity to provide information to resolve the issue, and issue a written determination.[21] If an identity issue arises after a worker has received a UI benefit payment, states must still continue paying benefits unless they have proper evidence that substantiates a reasonable basis to think fraud is occurring. If they do have proper evidence, they may pause payment pending completion of the ID verification process.[22]

States must also act promptly to verify an individual’s ID to comply with the federal requirement that states have methods of administration reasonably calculated to ensure full payment of UI benefits “when due.”[23]  This means that eligible workers must receive their first UI benefit payment as soon as is administratively feasible.[24]

Policy Recommendations

Eligible workers must have the ability to access UI benefits without facing excessive obstacles created by ID verification systems. The following recommendations are intended to facilitate access and equity and protect due process and privacy. In most cases, these recommendations can be implemented by state agencies under their regulatory authority, however state legislatures can also mandate that certain identity verification procedures be followed as the state of New Jersey did in 2022.[25]

States should:

1. Start with a risk assessment to determine which claims should be subject to ID verification. USDOL requires states to use cross-matches and data analytics to identify suspicious activity that may be indicative of fraud.[26] Agencies should work to ensure that fraud analytics do not have a disparate impact on workers of color and other historically underserved populations. Agencies should assume that data may be inaccurate and not automatically treat mismatched data as evidence of fraud. After a risk assessment, states can then require ID verification for claims identified as having the greatest risks.
2. Ensure workers’ due process rights are respected in the ID verification process.
   • Enable workers applying for UI benefits to complete their application even if ID verification is still pending. Payment of benefits can be delayed until ID verification is completed, but workers must receive a plain language notice explaining why their payment is pending, and what steps they must take to get their payment released. When these procedures take more than 30 days, workers must get expedited adjudication.
   • When a claim is denied because of failure to verify a worker’s identity, states must ensure an expedited, easily accessible appeals process. New Jersey passed state legislation to ensure workers can swiftly appeal if they are denied a digital identity credential.[27]
   • Ensure that completing an ID verification process in another state is not grounds to deny a claim.
   • Test ID verification systems against a valid name list to ensure applicants can use their real names.[28]
   • Ensure all worker-facing ID verification systems and communications are written in “plain language,” user tested, and compliant with regulations on language access and access for people with disabilities.[29]
3. Ensure that a state’s list of acceptable documents to verify ID includes a sufficient variety to provide for equal access to UI benefits for all workers as required by USDOL guidance.[30] For example, if the state requires a state driver’s license as proof for ID verification, it must offer a reasonable alternative for those who may not have that type of ID or provide sufficient time for workers to acquire other ID.
4. Where possible, enable workers to “reuse” the identity verification process they already underwent to get credentials such as a driver’s license rather than requiring a redundant new process.
5. Avoid the use of knowledge-based verification systems, as they can block workers with limited credit history and may be based on credit reports that are themselves inaccurate. The Government Accountability Office has also warned about the use of knowledge-based verification because data breaches have leaked the answers to many knowledge-based questions about individuals, making them no longer secret.[31]
6. Ensure all systems are completely mobile responsive since cell phones are the primary source of internet access for many workers.
7. Conduct a root cause analysis to identify the source of instances where workers legitimately applying for UI benefits have been harmed as a result of ID verification systems. As recommended by USDOL, states should seek to identify issues that may exist at larger scale and should make systemic modifications to the state’s ID verification processes as needed.[32]
   • In order to inform a root cause analysis, state agencies should gather anonymized individual level data on the ID verification process, such as the number and demographics of workers successful in verifying their identity, the number and demographics of workers denied because of being unable to verify identity, and the number and demographics of workers with payment pending because ID was not yet verified.
8. Ensure that workers applying for UI have an in-person, non-digital alternative, as required by USDOL guidance,[33] if they are unable or choose not to complete online identity verification. The in-person alternative should be an option but must never be required. New Jersey’s 2022 law on alternatives to digital identity credentials is one positive model.[34] In general, in-person, non-digital ID verification services should:
   • Have a reasonable wait time, call back, or appointment scheduling feature.
   • Be staffed by government merit staff.
   • Be located in accessible locations (such as USPS offices or job centers) and be accessible by public transportation to the greatest extent possible.
   • Include access to interpretation services to assist workers with limited English proficiency.
   • Provide documentation to individuals clearly stating whether their identity has been verified, and what next steps they must take, if any.
   • Allow expired IDs for in-person verification.
9. Ensure workers’ privacy is protected to the greatest extent possible.
   • Minimize the amount of data which is collected, used, and shared.
   • Do not retain biometric data after a worker’s identity has been verified.
   • Delete any documents submitted as part of the verification process after 30 days.
   • Ensure that biometric data is not used for any additional purpose beyond ID verification for UI benefits.
   • Disclose any use of biometric data at the beginning of the verification process and allow workers to opt out of the use of facial recognition or other biometric data collection without endangering access to UI benefits (see recommendation 8 about alternative means of identity verification).
10. Mandate vendor accountability and transparency. All ID verification system vendors must be required to meet the same standards for language and disability access, plain language, user testing, mobile phone accessibility, non-digital in-person alternatives for ID verification, appeal rights, and privacy protections as state agencies described above. Vendors must not share or sell any data collected as part of the ID verification process and must not use data for any purpose other than that for which it was collected. All elements of the product and steps of vendor’s process should be audited in real-world conditions by an independent party; found to be accurate, accessible, and to not have a disparate impact on the basis of race, gender, or other protected characteristics; and audits must be made publicly available. States should also ensure they have timely access to all relevant data around the vendor’s ID verification processes, including success and failure rates.

Research Findings and Arguments to Support Reform

Workers should not face barriers to accessing UI benefits because of the actions of international criminal enterprises.

A decades-long failure to invest in and reform the UI system made UI benefits an appealing target for organized crime during the COVID-19 pandemic. Chronically underfunded and under-staffed state UI systems were tasked with urgently implementing a major new set of federal pandemic UI programs even as the current UI system is woefully insufficient to address the needs of workers, particularly during an economic downturn. This provided multiple points of vulnerability for criminal enterprises to exploit.

The spike in fraudulent UI claims during the pandemic was primarily the result of criminal enterprises engaging in identity fraud, not of individual workers making improper claims.[35] As USDOL officials have explained, sophisticated organized crime rings that had previously stolen the identities of workers during private sector failures such as the Equifax data breach of 2017 seized upon the pandemic as a time to use those stolen identities to obtain pandemic UI benefits, which had a much higher value than regular UI benefits due to temporary supplements.[36]

Generally, fraudulent claims in regular state UI programs are relatively uncommon.[37] Therefore, states should be cautious in adopting ID verification systems that can block workers from claiming benefits they are entitled to and should take a risk-based approach to determining which claims should be subject to ID verification. The focus on stopping identity fraud can cause workers seeking UI benefits to be victimized twice: Workers are victimized first by the criminal enterprises that have stolen their identities and filed false UI claims in their names and a second time when they are caught in a snare intended to catch criminals and their benefits are delayed or denied entirely.

ID verification procedures can create additional barriers for workers of color, immigrant workers, and workers with disabilities.

ID verification systems can become yet another hurdle in an already-exclusionary UI system, in which Black workers, for example, are already 24 percent less likely to apply for or receive UI benefits than similarly situated white workers.[38] Since there is no perfect system for equitable ID verification, policymakers must be mindful of adopting any system: State agencies must conduct a risk analysis beforehand to determine if ID verification is necessary, provide non-digital alternatives for verifying identity, and collect data on the impact of ID verification in order to do a root cause analysis to understand the system’s harms and how to mitigate them.

• Unequal access to digital technology means online ID verification will exclude some workers.

Using an online ID verification system requires hardware (such as a smartphone or computer with a webcam), reliable internet access, and familiarity interacting with online systems. Yet Black and Latinx adults remain less likely than white adults to have a broadband internet connection at home.[39] Residents of rural areas are also less likely to have home broadband.[40] Not all ID verification systems are optimized for a smartphone, even as 15 percent of U.S. adults, including a quarter of Latinx adults, rely on a smartphone as their primary or only means to access the internet.[41] ID verification systems may not work with assistive technology used by people with disabilities.

• Facial recognition systems often do not perform equitably.

In March 2023, the USDOL’s Office of Inspector General released a memo warning about urgent equity and security concerns relating to the use of facial recognition technology in unemployment insurance programs.[42] The concern was based on numerous studies which find that facial recognition technology has error rates that vary dramatically across race and gender and could produce discriminatory outcomes for women and people of color.[43] For example, a 2019 study by the National Institute of Standards and Technology (NIST) found that facial recognition technology can produce racially and gender biased outcomes, including “false negative” errors that would incorrectly identify eligible workers applying for UI benefits as not being who they claim to be.[44]

• Knowledge-based verification systems may assume everyone has a credit history.

Knowledge-based verification systems ask workers personal questions that only they are presumed to know, such as details about cars they own, loans they have taken out, or past addresses. The questions are automatically generated based on data from credit bureaus. In addition to concerns about inaccurate credit reports and leaked information that affect all workers, these questions can create particular barriers for people with limited or no credit history,[45] who are disproportionately likely to be Black, Latinx, or to live in low-income communities.[46]

• Presenting identity documents can be a barrier.

A recent study found that 29 million U.S. adults did not have a current driver’s license and more than 7 million had no form of current government issued photo ID.[47] Workers of color, low-paid workers, and formerly incarcerated workers were less likely to have ID documents. Enabling workers to show a variety of ID documents, giving them time to obtain documents, and accepting expired documents can help to increase access.

• ID verification systems often fail to provide language access.

Workers with limited English proficiency have long faced barriers to accessing UI benefits, and the introduction of ID verification systems from outside vendors further compounds these hurdles. For example, researchers with the National Center for Law and Economic Justice documented the failure of New York’s ID verification contractor ID.me to translate key features, notices, or help guides into the state’s top languages, with long wait times for language assistance.[48]

• Additional concerns about equity and access in ID verification.

A wide range of biases are built into ID verification systems that can exclude eligible workers. For example, many systems were built exclusively with conventionally English names in mind, so that workers with non-English names may be rejected or flagged as potentially fraudulent simply for using a name that is considered by the system to be too long, too short, or otherwise differs from English naming conventions.[49] Similarly, ID verification systems may have a low threshold for the number of claims filed from the same address—failing to consider multigenerational households, homeless shelters, over even apartment buildings.[50] User-testing and analyzing ID verification systems is important because not all biases in the system are readily apparent.

ID verification procedures raise significant due process concerns.

When workers are blocked from applying for benefits because they cannot verify their identity, they are effectively denied benefits without sufficient notice or opportunity to appeal, a clear violation of federal due process rights.[51]

During the pandemic, changes to ID verification standards were frequently made without notice. Agencies expressed concern that revealing the standards or providing advance notification would alert fraudsters. However, if workers are required to prove their identity using a set of standards, then they need to understand what the requirements are. Unclear notices are also a denial of due process rights: An inability to verify identity is often described with a blanket statement such as “failure to comply with agency procedures.” This overly general reason for denial prevents workers from being able to meaningfully contest the denial, bring their claim into compliance, and access benefits.

ID verification procedures raise privacy concerns.

ID verification procedures raise a range of privacy concerns for workers applying for UI. Increased use of facial recognition technology in UI applications is especially troubling because facts about the unique characteristics of a worker’s facial structure or other biometric data are permanent and cannot be changed. Therefore, it is vital that sensitive identity data be protected, not retained beyond the immediate need, and not be used for anything other than the stated purpose of verifying a worker’s identity for purposes of accessing UI benefits. All workers must also have a clear option to refuse the use of facial recognition technology and verify their identity in another way.

The use of private ID verification system vendors raises additional concerns that data could be sold to other businesses, stored insecurely, or otherwise misused. In 2023, the USDOL Office of the Inspector General assessed contracts between 24 state workforce agencies and ID verification service contractors that used facial recognition technology.[52] The agency found that a majority of ID verification contracts did not address requirements for data storage or procedures for destroying or disposing of data collected. Many had no indication of the identity standards for facial recognition.

Government retention of biometric data can also cause alarm that publicly managed databases that were created to facilitate access to benefits may be repurposed for use by law enforcement.[53] Concerns that the government may misuse data provided as part of the ID verification process can undermine trust and deter workers from applying for UI benefits.

Data and State Comparison Resources

Learn what online ID verification procedures your state is using.

The Beeck Center’s resource on Digital Authentication and Identity Proofing in UI Applications documents digital authentication and identity proofing steps for online unemployment insurance applications in each state, based on publicly available information. The resource includes information about what forms of ID documents each state requires, whether workers have a choice of ID proofing methods, and screenshots of relevant ID verification webpages as well as other information. It is updated periodically.

Evaluate private vendors your state uses or is considering for ID verification.

Consult the Identity Proofing Vendor Comparison from U.S. Digital Response (USDR) for an overview of the leading ID verification vendors. The resource includes key considerations for each vendor, including what methods it uses to verify identity, what the user experience is like during identity proofing, and other factors.

Investigate whether ID verification procedures might be increasing UI denials in your state.

USDOL does not currently collect or report data on the number of workers denied UI benefits as a result of an inability to verify their identity. In some states, ID verification occurs before workers can even file a claim, so workers who cannot verify their identity are not counted as denied claims by the UI system. However, if workers file a claim but are later denied due to inability to verify their identity, they would be considered nonseparation denials—meaning they were denied benefits for a reason unrelated to their separation from the most recent employer.

You can view national or state data for denials by using the Century Foundation’s Unemployment Insurance Data Explorer. This tool utilizes USDOL data in a more user-friendly format. Select your state and view data on “Non-Separation Denial Breakdown” and “Non-Separation Denial Rates.” Claims denied because of incomplete ID verification procedures should appear as “Reporting/Call-ins/Etc.” (if the worker didn’t respond to the agency’s ID verification request) or “Other” (if the worker responded but didn’t successfully verify their identity).[54] Note that ID verification is not the only reason for denial in either of these categories—USDOL reports that they are creating a new category for ID verification issues.

References and Essential Articles

Relevant UI Program Letters

No. 16-21 (Identity Verification for Unemployment Insurance (UI) Claims) https://www.dol.gov/agencies/eta/advisories/unemployment-insurance-program-letter-no-16-21.

No. 11-23 (Announcement of Grant Opportunities and National Identity (ID) Verification Offering under the American Rescue Plan Act (ARPA)) https://www.dol.gov/agencies/eta/advisories/uipl-11-23.

Identity verification for unemployment insurance

Elizabeth Bynum Sorrell and Ariel Kennan, Digital Authentication and Identity Proofing in Unemployment Insurance (UI) Applications, Digital Benefits Network at the Beeck Center for Social Impact + Innovation at Georgetown University, May 19, 2023, https://www.digitalbenefitshub.org/digital-authentication-and-identity-proofing-data/unemployment-insurance.

Alyssa Levitz, Waldo Jaquith, and Michael Smedberg, Identity Proofing for UI Agencies, U.S. Digital Response, May 2021, https://usdr.gitbook.io/unemployment-insurance-modernization/identity-proofing-vendor-comparison/identity-proofing-vendor-comparison.

A Playbook for Improving Unemployment Insurance Delivery: Identity Verification, New America, 2022,  https://improveunemployment.com/identity_verification/.

Carolyn R. Hantz, Alert Memorandum: ETA and States Need to Ensure the Use of Identity Verification Service Contractors Results in Equitable Access to UI Benefits and Secure Biometric Data, U.S. Department of Labor Office of the Inspector General, March 31, 2023, https://www.oig.dol.gov/public/reports/oa/2023/19-23-005-03-315.pdf.

General resources on identity verification and digital identity for public benefits

Digital Identity resource page, Digital Benefits Network at the Beeck Center for Social Impact + Innovation at Georgetown University, https://www.digitalbenefitshub.org/digital-identity.

Digital Identity Glossary, Digital Benefits Network at the Beeck Center for Social Impact + Innovation at Georgetown University, December 2022,  https://www.digitalbenefitshub.org/resources/digital-identity-glossary.

Paul A. Grassi, Michael E. Garcia, and James L. Fenton, Digital Identity Guidelines, NIST Special Publication 800-63-3, National Institute of Standards and Technology, June 2017, https://pages.nist.gov/800-63-3/.

Michael Yang and Elizabeth Laird, Digital Identity Verification: Best Practices for Public Agencies, Center for Democracy and Technology, October 2022, https://cdt.org/insights/digital-identity-verification-best-practices-for-public-agencies/.

Equity considerations in identity verification

Alyssa Levitz, Race and Inequity in Identity Proofing Methods, U.S. Digital Response, February 2, 2021, https://usdr.gitbook.io/unemployment-insurance-modernization/identity-proofing-vendor-comparison/race-and-inequity-in-identity-proofing-methods.

Designed to Exclude: New York’s Failure to Provide Compensation and Language Access to Unemployed Workers, National Center for Law and Economic Justice, August 24, 2022, https://nclej.org/news/designed-to-exclude.

Specific identity verification vendors and technologies

Joy Buolamwini, Vicente Ordóñez, Jamie Morgenstern, and Erik Learned-Miller, Facial Recognition Technologies: A Primer, Algorithmic Justice League, May 29, 2020, https://www.ajl.org/federal-office-call.

Sharon M. Dietrich, ID.me Presents Barriers for Low Income People Seeking Unemployment Insurance and Other Government Benefits, Community Legal Services of Philadelphia, November 2021, https://clsphila.org/wp-content/uploads/2021/11/IDme-issue-brief-final-11-2-2021.pdf.

Potential Disparate Impacts of ID.me for Unemployment Insurance Claimants in Oregon – Anonymized, State of Oregon Employment Department, February 18, 2022, https://unemployment.oregon.gov/uploads/docs/2022.02.18-Potential-ID.Me-Disparate-Impacts-FINAL.pdf.

Jennifer Wagner and Genevieve Gaudet, Removing Barriers to Access From Remote Identity Proofing, Center on Budget and Policy Priorities, April 22, 2020, https://www.cbpp.org/research/health/removing-barriers-to-access-from-remote-identity-proofing.

The Role of the Postal Service in Identity Verification, Office of Inspector General, U.S. Postal Service, May 11, 2022, https://www.uspsoig.gov/reports/white-papers/role-postal-service-identity-verification.